Variational inference
Fit a hierarchical count model with brm_svi_fit, compare its Gaussian approximation with NUTS, continue the fit as new observations arrive, and combine fits on slices into a starting point. This example shows where the approximation loses population uncertainty even when its site-effect estimates agree with NUTS.
The model and the data
The synthetic data represent monitoring sites visited repeatedly. At each visit a count is recorded together with a covariate x, and the count rate of a site depends on x through the site's own intercept and slope. The two site effects are correlated. The data are a NamedTuple of nested vectors, one entry per site holding that site's visits; no table with one row per observation is needed.
# The values the data are simulated from.
const GENERATING = (; a=1.0, b=0.4, sd_a=0.5, sd_b=0.25, cor_ab=0.3)
# `n_sites` sites, each visited `visits` times. Each site has its own intercept and
# slope, drawn together from a bivariate normal around the population values; the
# count at a visit is Poisson with log rate `a_i + b_i * x`. The data are a NamedTuple
# of nested vectors: one entry per site, holding that site's visits.
function simulate_sites(; n_sites=400, visits=8, seed=1, g=GENERATING)
rng = Xoshiro(seed)
C = [g.sd_a^2 g.cor_ab * g.sd_a * g.sd_b; g.cor_ab * g.sd_a * g.sd_b g.sd_b^2]
L = cholesky(C).L
effects = [L * randn(rng, 2) for _ in 1:n_sites]
x = [randn(rng, visits) for _ in 1:n_sites]
y = [[rand(rng, Poisson(exp(g.a + e[1] + (g.b + e[2]) * xj))) for xj in xi]
for (e, xi) in zip(effects, x)]
data = (; site=[string("s", i) for i in 1:n_sites], x, y)
(; data, effects)
endThe model has a @plate over sites. The population part, a ~ 1 + (1 | s | site) and b ~ 1 + (1 | s | site), gives each site an intercept and a slope drawn from one correlated block s. The plate cell reads that site's two values and observes its counts.
Correlated site intercepts and slopes for Poisson counts
panes: brm|stanblocks|rkppl|stan|jax:3function site_counts_model(data = site_counts_data())
@brm data begin
a ~ 1 + (1 | s | site)
b ~ 1 + (1 | s | site)
effect(a, Intercept) ~ Normal(0, 1)
effect(b, Intercept) ~ Normal(0, 1)
sd(:, s) ~ Exponential(1)
@plate for i in eachindex(a)
y[i] ~ Poisson(exp.(a[i] .+ b[i] .* x[i]))
end
end
end
# ═══════════════════════════════════════════════════════════════════════════════
# 3. Fitting (not evaluated by the docs build; needs the test environment)
# ═══════════════════════════════════════════════════════════════════════════════SBBRMI with data keys = [:b_s_site_lkj_eta, :kernel_nsub_y_cell, :n_site, :n_terms_s_site, :site_idx, :x, :y]
configured submodels:
ranef_correlated_draws_generic_configured_1 = Base.merge(BayesianRegressionModels.ranef_correlated_draws_generic, quote
tau ~ (StanBlocks.builtin.restricted)(exponential, 1.0 ./ 1; n = n_terms, lower = 0.0)
end)
emitted @slic body:
begin
b_s_site ~ ranef_correlated_draws_generic_configured_1(; group_idx = site_idx, n_groups = n_site, n_terms = n_terms_s_site, lkj_eta = b_s_site_lkj_eta)
X_a = hcat(rep_vector(1.0, num_elements(site_idx)))
pop_a ~ _popefs_normal(; X = X_a, beta_loc = [0], beta_scale = [1])
r_a_s_site = b_s_site[site_idx, 1]
a = pop_a + r_a_s_site
X_b = hcat(rep_vector(1.0, num_elements(site_idx)))
pop_b ~ _popefs_normal(; X = X_b, beta_loc = [0], beta_scale = [1])
r_b_s_site = b_s_site[site_idx, 2]
b = pop_b + r_b_s_site
y_cell ~ plate(a, y, b, x; outer = (kernel_nsub_y_cell,)) do a, y, b, x
#= brm-docs-example.jl:9 =#
y ~ poisson(exp.(a .+ b .* x))
#= brm-docs-example.jl:10 =#
nothing
end
end"""
Group-level (varying) effects of grouping `g`, `K` correlated margins per level.
Non-centered, from the scales `tau[1:K]` (drawn from their priors), the Cholesky factor of the margins' correlation `L ~ LKJCholesky(K, eta)` and standardized effects
`z[1:length(levels(g)), 1:K] ~ Normal(0, 1)`. Returns the
level-by-margin matrix `z * transpose(tau .* L)`, one row per level of `g` in order.
"""
brm_correlated_group_effects(g, K, eta) = begin
tau[1:K] .~ Exponential.(1)
L ~ LKJCholesky(K, eta)
z[1:length(levels(g)), 1:K] .~ Normal.(0, 1)
return z * transpose(tau .* L)
end
begin
site_fit_levels = brm_fit_levels(site_fit)
b_s_site ~ brm_correlated_group_effects(site_fit_levels, 2, 1.0)
site_level = brm_level_indices(site, site_fit_levels)
ranef_a_s_site = b_s_site[site_level, 1]
ranef_b_s_site = b_s_site[site_level, 2]
a_Intercept ~ Normal(0.0, 1.0)
a = a_Intercept .+ ranef_a_s_site
b_Intercept ~ Normal(0.0, 1.0)
b = b_Intercept .+ ranef_b_s_site
@plate for i = eachindex(x)
y_cell_argument_y_1[i] = exp.(a[i] .+ b[i] .* x[i])
end
@plate for i = eachindex(y)
y[i] .~ Poisson.(y_cell_argument_y_1[i])
end
endfunctions {
matrix hcat(vector x) {
int n = dims(x)[1];
return to_matrix(x, n, 1);
}
vector poisson_lpmfs(
array[] int obs,
vector lambda
) {
return jbroadcasted_poisson_lpmfs(obs, lambda);
}
vector jbroadcasted_poisson_lpmfs(
array[] int x1,
vector x2
) {
int n = dims(x1)[1];
vector[n] rv;
for(i in 1:n) {
rv[i] = poisson_lpmfs(broadcasted_getindex(x1, i), broadcasted_getindex(x2, i));
}
return rv;
}
real poisson_lpmfs(int args1, real args2) {
return poisson_lpmf(args1 | args2);
}
int broadcasted_getindex(array[] int x, int i) {
return x[i];
}
real broadcasted_getindex(vector x, int i) {
return x[i];
}
array[] int poisson_int_rng(
int anontok__1,
vector p
) {
int n = anontok__1;
return poisson_rng(p);
}
int ragged_end(array[] int ends, int i) {
return ends[i];
}
int ragged_start(
array[] int ends,
int i
) {
if((i == 1)) {
return 1;
} else {
return (1 + ends[(i - 1)]);
}
}
vector jbroadcasted_exp(
vector x1
) {
int n = dims(x1)[1];
vector[n] rv;
for(i in 1:n) {
rv[i] = exp(broadcasted_getindex(x1, i));
}
return rv;
}
}
data {
int n_terms_s_site;
real b_s_site_lkj_eta;
int n_site;
int site_idx_n;
array[site_idx_n] int site_idx;
int y_mem_n;
int y_ends_n;
tuple(array[y_mem_n] int, array[y_ends_n] int) y;
int kernel_nsub_y_cell;
int x_ends_n;
int x_mem_n;
tuple(vector[x_mem_n], array[x_ends_n] int) x;
}
transformed data {
matrix[num_elements(site_idx), 1] X_a = hcat(rep_vector(1.0, num_elements(site_idx)));
int pop_a_n_covariates = 1;
matrix[num_elements(site_idx), 1] X_b = hcat(rep_vector(1.0, num_elements(site_idx)));
int pop_b_n_covariates = 1;
}
parameters {
cholesky_factor_corr[n_terms_s_site] b_s_site_L;
vector<lower=0.0>[n_terms_s_site] b_s_site_tau;
vector[(n_terms_s_site * n_site)] b_s_site_z_flat;
vector[pop_a_n_covariates] pop_a_beta_pop;
vector[pop_b_n_covariates] pop_b_beta_pop;
}
transformed parameters {
matrix[n_terms_s_site, n_site] b_s_site_z = to_matrix(b_s_site_z_flat, n_terms_s_site, n_site);
matrix[n_site, n_terms_s_site] b_s_site = ((diag_pre_multiply(b_s_site_tau, b_s_site_L) * b_s_site_z)');
vector[num_elements(site_idx)] pop_a = (X_a * pop_a_beta_pop);
vector[site_idx_n] r_a_s_site = b_s_site[site_idx, 1];
vector[num_elements(site_idx)] a = (pop_a + r_a_s_site);
vector[num_elements(site_idx)] pop_b = (X_b * pop_b_beta_pop);
vector[site_idx_n] r_b_s_site = b_s_site[site_idx, 2];
vector[num_elements(site_idx)] b = (pop_b + r_b_s_site);
}
model {
b_s_site_L ~ lkj_corr_cholesky(b_s_site_lkj_eta);
b_s_site_tau ~ exponential((1.0 ./ 1));
b_s_site_z_flat ~ std_normal();
pop_a_beta_pop ~ normal([0]', [1]');
pop_b_beta_pop ~ normal([0]', [1]');
for(plate_i__pl_1 in 1:kernel_nsub_y_cell) {
y.1[ragged_start(y.2, plate_i__pl_1):ragged_end(y.2, plate_i__pl_1)] ~ poisson(
jbroadcasted_exp(
(
a[plate_i__pl_1] +
(b[plate_i__pl_1] .* x.1[ragged_start(x.2, plate_i__pl_1):ragged_end(x.2, plate_i__pl_1)])
)
)
);
}
}
generated quantities {
array[num_elements(y.1)] int y_gen;
vector[num_elements(y.2)] y_likelihood;
for(plate_i__pl_1 in 1:kernel_nsub_y_cell) {
y_gen[ragged_start(y.2, plate_i__pl_1):ragged_end(y.2, plate_i__pl_1)] = poisson_int_rng(
(1 + (ragged_end(y.2, plate_i__pl_1) - ragged_start(y.2, plate_i__pl_1))),
jbroadcasted_exp(
(
a[plate_i__pl_1] +
(b[plate_i__pl_1] .* x.1[ragged_start(x.2, plate_i__pl_1):ragged_end(x.2, plate_i__pl_1)])
)
)
);
y_likelihood[plate_i__pl_1] = poisson_lpmf(y.1[ragged_start(y.2, plate_i__pl_1):ragged_end(y.2, plate_i__pl_1)] |
jbroadcasted_exp(
(
a[plate_i__pl_1] +
(b[plate_i__pl_1] .* x.1[ragged_start(x.2, plate_i__pl_1):ragged_end(x.2, plate_i__pl_1)])
)
)
);
}
}using ReactiveKernels, ReactiveKernelsPPL
rk = RKBRMI(model)
source = jax_source(rk.model, rk_bound_plan(rk))
write("model.py", source)
inputs = rk_bound_data(rk, source) # exactly the inputs prepare_data readsimport jax
import model as m
prepared = m.prepare_data(**inputs)
lp = m.log_density(u, **prepared) # u: unconstrained, length m.dimension(prepared)
lp = jax.jit(m.log_density)(u, **prepared) # one compiled function for every dataset of these shapes"""Generated by ReactiveKernelsPPL.jax_source from a built @rkppl program; do not edit.
Data: site, site_fit, x, y.
Model documentation:
- `b_s_site`: Group-level (varying) effects of grouping `g`, `K` correlated margins per level.
Non-centered, from the scales `tau[1:K]` (drawn from their priors), the Cholesky factor of the margins' correlation `L ~ LKJCholesky(K, eta)` and standardized effects
`z[1:length(levels(g)), 1:K] ~ Normal(0, 1)`. Returns the
level-by-margin matrix `z * transpose(tau .* L)`, one row per level of `g` in order.
"""
import numpy as np
import scipy.special as ssp
from typing import Any
import jax
import jax.numpy as jnp
import jax.scipy.linalg
import jax.scipy.special as jsp
from jax.typing import ArrayLike
jax.config.update("jax_enable_x64", True)
DATA = {"site": {"dtype": "object", "ndim": 1, "ragged": False}, "site_fit": {"dtype": "object", "ndim": 1, "ragged": False}, "x": {"dtype": "float64", "ndim": 1, "ragged": True}, "y": {"dtype": "int64", "ndim": 1, "ragged": True}}
INPUTS = ("unconstrained", "ppl_last_b_s_site_z", "ppl_axis1_b_s_site_z", "site_level", "plate_argument", "x", "x_lengths", "x_flat", "y", "y_lengths", "y_flat")
PREPARED = {"ppl_last_b_s_site_z": {"dtype": "int64", "ndim": 0, "ragged": False}, "ppl_axis1_b_s_site_z": {"dtype": "int64", "ndim": 0, "ragged": False}, "site_level": {"dtype": "int64", "ndim": 1, "ragged": False}, "plate_argument": {"dtype": "int64", "ndim": 1, "ragged": False}, "x": {"dtype": "float64", "ndim": 1, "ragged": True}, "y": {"dtype": "int64", "ndim": 1, "ragged": True}}
UNTRANSLATED = {}
def _rk_order_key_np(x):
# Julia's isless order of an array as sortable keys: a float is an integer
# key, -0.0 before 0.0 and every NaN last, equal for values isequal treats
# as equal. Integer and string arrays are their own keys.
x = np.asarray(x)
if x.dtype.kind != "f":
return x
bits = np.where(np.isnan(x), np.nan, x).astype(np.float64).view(np.int64)
return np.where(bits < 0, bits ^ np.int64(0x7FFFFFFFFFFFFFFF), bits)
def _rk_unique_np(x):
# Julia's unique of a vector: the first value of each isequal class, in
# order.
x = np.asarray(x)
_, first = np.unique(_rk_order_key_np(x), return_index=True)
return x[np.sort(first)]
def _rk_sort_np(x):
# Julia's sort of a vector: ascending in isless order, equal values in
# their order.
x = np.asarray(x)
return x[np.argsort(_rk_order_key_np(x), kind="stable")]
def _brm_sorted_labels_np(source):
labels = _rk_unique_np(source)
sorted_ = _rk_sort_np(labels)
return sorted_
def _brm_fit_levels_np(source):
value = _brm_sorted_labels_np(source)
levels = value
return levels
def _rk_dict_key(x):
if isinstance(x, tuple):
return (1, tuple(map(_rk_dict_key, x)))
if isinstance(x, np.ndarray):
return (2, x.shape, tuple(map(_rk_dict_key, x.ravel(order="F"))))
if isinstance(x, (complex, np.complexfloating)):
if x.imag == 0 and not np.signbit(x.imag):
return _rk_dict_key(x.real)
return (3, _rk_dict_key(x.real), _rk_dict_key(x.imag))
if isinstance(x, (float, np.floating)):
if np.isnan(x):
return (4,)
if x == 0 and np.signbit(x):
return (5,)
return (0, x)
class _RKDict:
def __init__(self, pairs=()):
self.data = {}
for key, value in pairs:
self[key] = value
def __len__(self):
return len(self.data)
def __contains__(self, key):
return _rk_dict_key(key) in self.data
def __getitem__(self, key):
return self.data[_rk_dict_key(key)][1]
def __setitem__(self, key, value):
self.data[_rk_dict_key(key)] = (key, value)
def get(self, key, default):
item = self.data.get(_rk_dict_key(key))
return default if item is None else item[1]
def setdefault(self, key, default):
return self.data.setdefault(_rk_dict_key(key), (key, default))[1]
def pop(self, key, default):
item = self.data.pop(_rk_dict_key(key), None)
return default if item is None else item[1]
def items(self):
return self.data.values()
def keys(self):
return (item[0] for item in self.items())
def values(self):
return (item[1] for item in self.items())
def copy(self):
return _RKDict(self.items())
def _brm_level_indices_np(labels, source):
value = _brm_fit_levels_np(source)
levels = value
lookup = _RKDict()
for k in range(1, len(levels) + 1):
lookup[levels[k - 1]] = k
indices = np.zeros((len(labels),), dtype=np.int64)
for r in range(1, len(labels) + 1):
position = lookup.get(labels[r - 1], 0)
if position == 0:
raise ValueError("".join([str("RK backend: group label `"), str(labels[r - 1]), str("` is not one of the "), str(len(levels)), str(" fitted levels. A replay keeps the fitted levels of every grouping it does not resample; name the grouping in `resample_groups` to take the new data's own levels.")]))
indices[r - 1] = position
return indices
def _rk_levels_np(x):
arr = np.asarray(x)
values = sorted(set(v for v in arr.tolist() if v is not None))
return np.array(values, dtype=object if arr.dtype == object else arr.dtype)
class StaticInt(int):
"""An integer of the prepared data that sets a shape.
It is an `int` everywhere, and `jax.jit` treats it as static, so the
functions jit with the prepared data as arguments; a new value traces
them again.
"""
def __repr__(self):
return "StaticInt(%d)" % int(self)
jax.tree_util.register_pytree_node(StaticInt, lambda s: ((), int(s)), lambda v, _: StaticInt(v))
def _rk_pad_rows(rows, dtype, rank=1):
# A ragged value as the functions take it: the padded array (each row
# repeats its last entry; an empty row is zeros), the lengths, and the
# positions of the valid entries in the flattened padded array, whose
# length is the total. `dtype` is the element type the program declares,
# so an empty row (`[]`) keeps an integer column integer.
rows = [np.asarray(r, dtype=dtype) for r in rows]
# One length per cell describes its first axis. Other axes retain a
# common shape; independently ragged axes need separate extents.
tail = rows[0].shape[1:] if rows else (1,) * (rank - 1)
if any(r.ndim != rank or r.shape[1:] != tail for r in rows):
raise ValueError("ragged cells need their declared rank and common trailing dimensions")
lengths = np.asarray([r.shape[0] for r in rows], dtype=np.int64)
width = max([1 if rank > 1 and not rows else 0] + [int(n) for n in lengths])
padded = np.stack([np.pad(r, ((0, width - r.shape[0]),) + ((0, 0),) * (rank - 1),
mode="edge") if r.shape[0] > 0
else np.zeros((width,) + tail, dtype=dtype) for r in rows]) if rows else np.zeros((0, width) + tail, dtype=dtype)
flat = np.flatnonzero(np.arange(width)[None, :] < lengths[:, None])
return padded, lengths, flat
def brm_correlated_group_effects_density(
ppl_last_z: ArrayLike,
tau: ArrayLike,
unconstrained: ArrayLike,
*,
outputs: tuple[str, ...],
) -> dict:
"""Group-level (varying) effects of grouping `g`, `K` correlated margins per level.
Non-centered, from the scales `tau[1:K]` (drawn from their priors), the Cholesky factor of the margins' correlation `L ~ LKJCholesky(K, eta)` and standardized effects
`z[1:length(levels(g)), 1:K] ~ Normal(0, 1)`. Returns the
level-by-margin matrix `z * transpose(tau .* L)`, one row per level of `g` in order.
"""
results = {}
if any(name in outputs for name in ("ppl_lkj_diagonal_L", "ppl_lkj_partials_L", "ppl_lkj_raw_L", "ppl_prior_L")):
ppl_lkj_raw_L = unconstrained[4:5]
if any(name in outputs for name in ("ppl_lkj_diagonal_L", "ppl_lkj_partials_L", "ppl_prior_L")):
ppl_lkj_partials_L = jnp.tanh(ppl_lkj_raw_L)
if any(name in outputs for name in ("ppl_lkj_diagonal_L", "ppl_prior_L")):
out = jnp.ones((2,), dtype=jnp.float64)
tuple_ = (1, 2, 1.0)
i = tuple_[0]
j = tuple_[1]
d = tuple_[2]
def loop_body(p, carry):
(d_2, i_2, j_2, out_2) = carry
d_3 = d_2 * jnp.sqrt(1 - ppl_lkj_partials_L[p - 1] ** 2)
def then_arm():
i_3 = i_2 + 1
return (None, jnp.asarray(d_3, dtype=jnp.float64), jnp.asarray(i_3, dtype=jnp.int64), jnp.asarray(j_2, dtype=jnp.int64), out_2)
def else_arm():
out_3 = out_2.at[j_2 - 1].set(d_3)
tuple__2 = (1, j_2 + 1, 1.0)
i_4 = tuple__2[0]
j_3 = tuple__2[1]
d_4 = tuple__2[2]
return (None, jnp.asarray(d_4, dtype=jnp.float64), jnp.asarray(i_4, dtype=jnp.int64), jnp.asarray(j_3, dtype=jnp.int64), out_3)
branch = jax.lax.cond(i_2 + 1 < j_2, then_arm, else_arm)
return (branch[1], branch[2], branch[3], branch[4])
loop_carry = jax.lax.fori_loop(1, 2, loop_body, (jnp.asarray(d, dtype=jnp.float64), jnp.asarray(i, dtype=jnp.int64), jnp.asarray(j, dtype=jnp.int64), out))
ppl_lkj_diagonal_L = loop_carry[3]
if any(name in outputs for name in ("ppl_arrflat_z", "ppl_prior_z", "ppl_pw_prior_z")):
ppl_arrflat_z = unconstrained[5:ppl_last_z]
if any(name in outputs for name in ("ppl_prior_tau", "ppl_pw_prior_tau")):
def ppl_pw_prior_tau_cell(tau_2):
log_scale = np.log(1.0)
logpdf = jax.lax.cond(tau_2 < 0, lambda: -jnp.inf, lambda: -log_scale - tau_2 / 1.0)
return logpdf
ppl_pw_prior_tau = (jnp.zeros(tau.shape[:1], dtype=jnp.float64) if 0 in tau.shape[:1] else jax.vmap(ppl_pw_prior_tau_cell, in_axes=(0,))(tau))
if any(name in outputs for name in ("ppl_prior_tau",)):
ppl_prior_tau = jnp.sum(ppl_pw_prior_tau)
if any(name in outputs for name in ("ppl_prior_L",)):
total = -0.6931471805599453
def loop_body_2(i_2, carry_2):
(total_2,) = carry_2
total_3 = total_2 + (2 - i_2) * jnp.log(ppl_lkj_diagonal_L[i_2 - 1])
return (total_3,)
loop_carry_2 = jax.lax.fori_loop(2, 3, loop_body_2, (jnp.asarray(total, dtype=jnp.float64),))
ppl_prior_L = loop_carry_2[0]
if any(name in outputs for name in ("ppl_prior_z", "ppl_pw_prior_z")):
def ppl_pw_prior_z_cell(z):
log_scale = np.log(1.0)
standardized = (z - 0.0) / jnp.where((z == 0.0 if 1.0 == 0 else False), 1.0, 1.0)
standard_logpdf = -0.5 * np.log(2 * np.pi) - 0.5 * standardized ** 2
logpdf = standard_logpdf - jnp.where((z != 0.0 if 1.0 == 0 else False), 0.0, log_scale)
return logpdf
ppl_pw_prior_z = (jnp.zeros(ppl_arrflat_z.shape[:1], dtype=jnp.float64) if 0 in ppl_arrflat_z.shape[:1] else jax.vmap(ppl_pw_prior_z_cell, in_axes=(0,))(ppl_arrflat_z))
if any(name in outputs for name in ("ppl_prior_z",)):
ppl_prior_z = jnp.sum(ppl_pw_prior_z)
if "ppl_arrflat_z" in outputs:
results["ppl_arrflat_z"] = ppl_arrflat_z
if "ppl_lkj_diagonal_L" in outputs:
results["ppl_lkj_diagonal_L"] = ppl_lkj_diagonal_L
if "ppl_lkj_partials_L" in outputs:
results["ppl_lkj_partials_L"] = ppl_lkj_partials_L
if "ppl_lkj_raw_L" in outputs:
results["ppl_lkj_raw_L"] = ppl_lkj_raw_L
if "ppl_prior_L" in outputs:
results["ppl_prior_L"] = ppl_prior_L
if "ppl_prior_tau" in outputs:
results["ppl_prior_tau"] = ppl_prior_tau
if "ppl_prior_z" in outputs:
results["ppl_prior_z"] = ppl_prior_z
if "ppl_pw_prior_tau" in outputs:
results["ppl_pw_prior_tau"] = ppl_pw_prior_tau
if "ppl_pw_prior_z" in outputs:
results["ppl_pw_prior_z"] = ppl_pw_prior_z
return results
def _ppl_lkj_lower(z: ArrayLike, diagonal: ArrayLike, K: ArrayLike, P: ArrayLike) -> jax.Array:
"""Compute `_ppl_lkj_lower` from its authored source."""
out = jnp.zeros((K, K), dtype=jnp.float64)
out_2 = out.at[0, 0].set(1.0)
tuple_ = (1, 2, 1.0)
i = tuple_[0]
j = tuple_[1]
w = tuple_[2]
def loop_body(p, carry):
(i_2, j_2, out_3, w_2) = carry
out_4 = out_3.at[j_2 - 1, i_2 - 1].set(z[p - 1] * w_2)
w_3 = w_2 * jnp.sqrt(1 - z[p - 1] ** 2)
def then_arm():
i_3 = i_2 + 1
return (None, jnp.asarray(i_3, dtype=jnp.int64), jnp.asarray(j_2, dtype=jnp.int64), out_4, jnp.asarray(w_3, dtype=jnp.float64))
def else_arm():
out_5 = out_4.at[j_2 - 1, j_2 - 1].set(diagonal[j_2 - 1])
tuple__2 = (1, j_2 + 1, 1.0)
i_4 = tuple__2[0]
j_3 = tuple__2[1]
w_4 = tuple__2[2]
return (None, jnp.asarray(i_4, dtype=jnp.int64), jnp.asarray(j_3, dtype=jnp.int64), out_5, jnp.asarray(w_4, dtype=jnp.float64))
branch = jax.lax.cond(i_2 + 1 < j_2, then_arm, else_arm)
return (branch[1], branch[2], branch[3], branch[4])
carry_init = (jnp.asarray(i, dtype=jnp.int64), jnp.asarray(j, dtype=jnp.int64), out_2, jnp.asarray(w, dtype=jnp.float64))
loop_carry = jax.lax.fori_loop(1, P + 1, loop_body, carry_init) if 1 <= P else carry_init
return loop_carry[2]
def brm_correlated_group_effects(
L: ArrayLike,
tau: ArrayLike,
z: ArrayLike,
*,
outputs: tuple[str, ...],
) -> dict:
"""Group-level (varying) effects of grouping `g`, `K` correlated margins per level.
Non-centered, from the scales `tau[1:K]` (drawn from their priors), the Cholesky factor of the margins' correlation `L ~ LKJCholesky(K, eta)` and standardized effects
`z[1:length(levels(g)), 1:K] ~ Normal(0, 1)`. Returns the
level-by-margin matrix `z * transpose(tau .* L)`, one row per level of `g` in order.
"""
results = {}
if any(name in outputs for name in ("value",)):
value = z @ jnp.transpose(tau[..., None] * L)
if "value" in outputs:
results["value"] = value
return results
@jax.custom_jvp
def _loggamma(x1):
return jsp.gammaln(x1)
def _loggamma_partials_1(x1):
return (jsp.gammaln(x1), jsp.digamma(x1))
def _loggamma_jvp(primals, tangents):
active = tuple(not isinstance(t, jax.custom_derivatives.SymbolicZero) for t in tangents)
mask = sum((1 << i) for i, a in enumerate(active) if a)
if mask == 0:
value = _loggamma(*primals)
return value, jnp.zeros_like(value)
if mask == 1:
values = _loggamma_partials_1(*primals)
return values[0], values[1] * tangents[0]
_loggamma.defjvp(_loggamma_jvp, symbolic_zeros=True)
def prepare_data(site: Any, site_fit: Any, x: list, y: list) -> dict:
"""The functions' data inputs, computed from the raw data once per dataset.
Pass the data by the names the model reads them; a list of arrays is a ragged
column. Call the other functions with the result: `f(u, **prepared)`.
"""
site = np.asarray(site, dtype=object)
site_fit = np.asarray(site_fit, dtype=object)
x = [np.asarray(a, dtype=np.float64) for a in x]
y = [np.asarray(a, dtype=np.int64) for a in y]
value = _brm_fit_levels_np(site_fit)
site_fit_levels = value
value_2 = _brm_level_indices_np(site, site_fit_levels)
site_level = np.asarray(value_2, dtype=np.int64)
ppl_axis1_b_s_site_z = len(_rk_levels_np(site_fit_levels))
ppl_last_b_s_site_z = 5 + 2 * ppl_axis1_b_s_site_z
plate_argument = np.arange(1, len(x) + 1)
x_padded, x_lengths, x_flat = _rk_pad_rows(x, np.float64, 1)
y_padded, y_lengths, y_flat = _rk_pad_rows(y, np.int64, 1)
return {"ppl_last_b_s_site_z": StaticInt(ppl_last_b_s_site_z), "ppl_axis1_b_s_site_z": StaticInt(ppl_axis1_b_s_site_z), "site_level": np.asarray(site_level, dtype=np.int64), "plate_argument": np.asarray(plate_argument, dtype=np.int64), "x": x_padded, "x_lengths": x_lengths, "x_flat": x_flat, "y": y_padded, "y_lengths": y_lengths, "y_flat": y_flat}
def model(_entry: str, unconstrained: ArrayLike, **prepared: ArrayLike) -> Any:
"""Shared model computation. The generated entry wrappers select a static output."""
ppl_last_b_s_site_z = prepared["ppl_last_b_s_site_z"]
ppl_axis1_b_s_site_z = prepared["ppl_axis1_b_s_site_z"]
site_level = prepared["site_level"]
plate_argument = prepared["plate_argument"]
x = prepared["x"]
x_lengths = prepared["x_lengths"]
x_flat = prepared["x_flat"]
y = prepared["y"]
y_lengths = prepared["y_lengths"]
y_flat = prepared["y_flat"]
unconstrained = jnp.asarray(unconstrained)
site_level = jnp.asarray(site_level)
plate_argument = jnp.asarray(plate_argument)
x = jnp.asarray(x)
x_lengths = jnp.asarray(x_lengths)
x_flat = jnp.asarray(x_flat)
y = jnp.asarray(y)
y_lengths = jnp.asarray(y_lengths)
y_flat = jnp.asarray(y_flat)
if _entry in ("log_density", "log_likelihood", "constrain"):
a_Intercept = unconstrained[0]
b_Intercept = unconstrained[1]
if _entry in ("log_density", "log_prior", "log_likelihood", "constrain"):
plate_argument_2 = unconstrained[2:4]
def b_s_site_tau_cell(ppl_cell_b_s_site_tau):
constrain_2 = jnp.exp(ppl_cell_b_s_site_tau)
return constrain_2
b_s_site_tau = (jnp.zeros(plate_argument_2.shape[:1], dtype=jnp.float64) if 0 in plate_argument_2.shape[:1] else jax.vmap(b_s_site_tau_cell, in_axes=(0,))(plate_argument_2))
if _entry in ("log_density", "log_jacobian"):
plate_argument_3 = unconstrained[2:4]
def ppl_ljcells_b_s_site_tau_cell(ppl_cell_b_s_site_tau):
return ppl_cell_b_s_site_tau
ppl_ljcells_b_s_site_tau = (jnp.zeros(plate_argument_3.shape[:1], dtype=jnp.float64) if 0 in plate_argument_3.shape[:1] else jax.vmap(ppl_ljcells_b_s_site_tau_cell, in_axes=(0,))(plate_argument_3))
if _entry in ("log_density", "log_prior"):
ppl_prior_parameters_normal_mu = jnp.asarray([0.0, 0.0], dtype=jnp.float64)
ppl_prior_parameters_normal_sd = jnp.asarray([1.0, 1.0], dtype=jnp.float64)
prefix_stop = 2
if not (prefix_stop <= unconstrained.size):
raise ValueError("BoundsError: prefix exceeds array length")
plate_argument_4 = unconstrained[:max(prefix_stop, 0)]
def ppl_prior_parameters_normal_pw_cell(value, location, scale):
log_scale = jnp.log(scale)
standardized = (value - location) / jnp.where(jnp.logical_and(scale == 0, value == location), 1.0, scale)
standard_logpdf = -0.5 * np.log(2 * np.pi) - 0.5 * standardized ** 2
logpdf = standard_logpdf - jnp.where(jnp.logical_and(scale == 0, value != location), 0.0, log_scale)
return logpdf
ppl_prior_parameters_normal_pw = (jnp.zeros(plate_argument_4.shape[:1], dtype=jnp.float64) if 0 in plate_argument_4.shape[:1] and ppl_prior_parameters_normal_mu.shape[:1] == plate_argument_4.shape[:1] and ppl_prior_parameters_normal_sd.shape[:1] == plate_argument_4.shape[:1] else jax.vmap(ppl_prior_parameters_normal_pw_cell, in_axes=(0, 0, 0))(plate_argument_4, ppl_prior_parameters_normal_mu, ppl_prior_parameters_normal_sd))
ppl_prior_parameters_normal = jnp.sum(ppl_prior_parameters_normal_pw)
if _entry in ("log_density", "log_jacobian"):
plate_fold = jnp.sum(ppl_ljcells_b_s_site_tau)
b_s_site = brm_correlated_group_effects_density(
ppl_last_z=(ppl_last_b_s_site_z if _entry in ("log_density", "log_prior", "log_likelihood", "constrain") else None),
tau=(b_s_site_tau if _entry in ("log_density", "log_prior") else None),
unconstrained=unconstrained,
outputs=tuple(
name for name, needed in (
("ppl_arrflat_z", ("log_density", "log_prior", "log_likelihood", "constrain")),
("ppl_lkj_diagonal_L", ("log_density", "log_prior", "log_likelihood", "constrain")),
("ppl_lkj_partials_L", ("log_density", "log_prior", "log_likelihood", "log_jacobian", "constrain")),
("ppl_lkj_raw_L", ("log_density", "log_prior", "log_likelihood", "log_jacobian", "constrain")),
("ppl_prior_L", ("log_density", "log_prior")),
("ppl_prior_tau", ("log_density", "log_prior")),
("ppl_prior_z", ("log_density", "log_prior")),
("ppl_pw_prior_tau", ("log_density", "log_prior")),
("ppl_pw_prior_z", ("log_density", "log_prior")),
)
if _entry in needed
),
)
if _entry in ("log_density", "log_likelihood", "constrain"):
ppl_lkj_lower = _ppl_lkj_lower(b_s_site["ppl_lkj_partials_L"], b_s_site["ppl_lkj_diagonal_L"], 2, 1)
b_s_site_L = ppl_lkj_lower
if _entry in ("log_density", "log_jacobian"):
total = 0.0
tuple_ = (1, 2)
i = tuple_[0]
j = tuple_[1]
def loop_body(p, carry):
(i_2, j_2, total_2) = carry
total_3 = total_2 + (j_2 - i_2 + 1) / 2 * jnp.log(1 - b_s_site["ppl_lkj_partials_L"][p - 1] ** 2)
tuple__2 = jax.lax.cond(i_2 + 1 < j_2, lambda: (i_2 + 1, j_2), lambda: (1, j_2 + 1))
i_3 = tuple__2[0]
j_3 = tuple__2[1]
return (i_3, j_3, total_3)
loop_carry = jax.lax.fori_loop(1, 2, loop_body, (jnp.asarray(i, dtype=jnp.int64), jnp.asarray(j, dtype=jnp.int64), jnp.asarray(total, dtype=jnp.float64)))
ppl_lkj_logjac_b_s_site_L = loop_carry[2]
if _entry in ("log_density", "log_likelihood", "constrain"):
b_s_site_z = jnp.reshape(b_s_site["ppl_arrflat_z"], (ppl_axis1_b_s_site_z, 2), order="F")
if _entry in ("log_density", "log_prior"):
prior = ppl_prior_parameters_normal + b_s_site["ppl_prior_tau"] + b_s_site["ppl_prior_L"] + b_s_site["ppl_prior_z"]
if _entry in ("log_density", "log_jacobian"):
log_jacobian_2 = plate_fold + ppl_lkj_logjac_b_s_site_L
if _entry in ("log_density", "log_likelihood"):
b_s_site_2 = brm_correlated_group_effects(
L=(b_s_site_L if _entry in ("log_density", "log_likelihood") else None),
tau=(b_s_site_tau if _entry in ("log_density", "log_likelihood") else None),
z=(b_s_site_z if _entry in ("log_density", "log_likelihood") else None),
outputs=tuple(
name for name, needed in (
("value", ("log_density", "log_likelihood")),
)
if _entry in needed
),
)
ranef_a_s_site = b_s_site_2["value"][site_level - 1, 0]
ranef_b_s_site = b_s_site_2["value"][site_level - 1, 1]
a = a_Intercept + ranef_a_s_site
b = b_Intercept + ranef_b_s_site
def y_cell_argument_y_1_cell(ppl_pi_y_cell_argument_y_1, x_2, x_n, a_2, b_2):
ppl_linear_x = x_2
plate_value = jnp.exp(a_2[ppl_pi_y_cell_argument_y_1 - 1] + b_2[ppl_pi_y_cell_argument_y_1 - 1] * ppl_linear_x[ppl_pi_y_cell_argument_y_1 - 1])
return (plate_value, x_n[ppl_pi_y_cell_argument_y_1 - 1])
plate_value, plate_lengths = ((jnp.zeros((0, 0), dtype=jnp.float64), jnp.zeros((0,), dtype=jnp.int64)) if 0 in plate_argument.shape[:1] else jax.vmap(y_cell_argument_y_1_cell, in_axes=(0, None, None, None, None))(plate_argument, x, x_lengths, a, b))
y_cell_argument_y_1 = plate_value
def ppl_pw_y_resp_totals_cell(ppl_group_y, ppl_group_y_n, ppl_group_y_cell_argument_y_1, ppl_group_y_cell_argument_y_1_n):
def ppl_pw_y_resp_group_cell(ppl_group_y_2, ppl_group_y_cell_argument_y_1_2, active_cell):
def then_arm_2():
def _integer_observation() -> jax.Array:
"""Compute `_integer_observation` from its authored source."""
return True
integer_observation = _integer_observation()
def then_arm():
def else_arm():
value = _loggamma(ppl_group_y_2 + 1.0)
return ppl_group_y_2 * jnp.log(ppl_group_y_cell_argument_y_1_2) - ppl_group_y_cell_argument_y_1_2 - value
return jax.lax.cond(ppl_group_y_2 == 0, lambda: -ppl_group_y_cell_argument_y_1_2, else_arm)
return jax.lax.cond((ppl_group_y_2 >= 0 if integer_observation else False), then_arm, lambda: -jnp.inf)
ppl_density = jax.lax.cond(ppl_group_y_cell_argument_y_1_2 >= 0, then_arm_2, lambda: -jnp.inf)
return ppl_density
ppl_pw_y_resp_group = (jnp.zeros(ppl_group_y.shape[:1], dtype=jnp.float64) if 0 in ppl_group_y.shape[:1] and ppl_group_y_cell_argument_y_1.shape[:1] == ppl_group_y.shape[:1] else jax.vmap(ppl_pw_y_resp_group_cell, in_axes=(0, 0, 0))(ppl_group_y, ppl_group_y_cell_argument_y_1, jnp.logical_and(jnp.broadcast_to((jnp.arange(ppl_group_y.shape[0]) < ppl_group_y_n), ppl_group_y.shape[:1]), jnp.broadcast_to((jnp.arange(ppl_group_y_cell_argument_y_1.shape[0]) < ppl_group_y_cell_argument_y_1_n), ppl_group_y_cell_argument_y_1.shape[:1]))))
ppl_group_total = jnp.sum(jnp.where((jnp.arange(ppl_pw_y_resp_group.shape[0]) < ppl_group_y_n), ppl_pw_y_resp_group, 0.0))
return ppl_group_total
ppl_pw_y_resp_totals = (jnp.zeros(y.shape[:1], dtype=jnp.float64) if 0 in y.shape[:1] and y_cell_argument_y_1.shape[:1] == y.shape[:1] else jax.vmap(ppl_pw_y_resp_totals_cell, in_axes=(0, 0, 0, 0))(y, y_lengths, y_cell_argument_y_1, x_lengths[plate_argument - 1]))
ppl_lik_y_resp = jnp.sum(ppl_pw_y_resp_totals)
if _entry in ("log_density",):
posterior = prior + ppl_lik_y_resp + log_jacobian_2
if not (unconstrained.size == ppl_last_b_s_site_z):
raise ValueError("[layout] packed length ≠ layout total: a packed point holds the coordinates of the binding whose layout packed it")
if _entry == "log_density":
return jnp.asarray(posterior, dtype=jnp.float64)
if _entry == "log_prior":
return jnp.asarray(prior, dtype=jnp.float64)
if _entry == "log_likelihood":
return jnp.asarray(ppl_lik_y_resp, dtype=jnp.float64)
if _entry == "log_jacobian":
return jnp.asarray(log_jacobian_2, dtype=jnp.float64)
if _entry == "constrain":
return {"a_Intercept": jnp.asarray(a_Intercept, dtype=jnp.float64), "b_Intercept": jnp.asarray(b_Intercept, dtype=jnp.float64), "b_s_site": {"tau": jnp.asarray(b_s_site_tau), "L": jnp.asarray(b_s_site_L), "z": jnp.asarray(b_s_site_z)}}
def log_density(unconstrained: ArrayLike, **prepared: ArrayLike) -> jax.Array:
"""The log density a sampler targets at `unconstrained`, the packed unconstrained parameters (length `dimension(prepared)`), for the data `prepare_data` returned: `f(u, **prepared)`.
The sum of `log_prior`, `log_likelihood` and `log_jacobian`.
"""
return model("log_density", unconstrained, **prepared)
def log_prior(unconstrained: ArrayLike, **prepared: ArrayLike) -> jax.Array:
"""The log prior density of the constrained parameters at `unconstrained`, the packed unconstrained parameters (length `dimension(prepared)`), for the data `prepare_data` returned: `f(u, **prepared)`."""
return model("log_prior", unconstrained, **prepared)
def log_likelihood(unconstrained: ArrayLike, **prepared: ArrayLike) -> jax.Array:
"""The log likelihood of the observations at `unconstrained`, the packed unconstrained parameters (length `dimension(prepared)`), for the data `prepare_data` returned: `f(u, **prepared)`."""
return model("log_likelihood", unconstrained, **prepared)
def log_jacobian(unconstrained: ArrayLike, **prepared: ArrayLike) -> jax.Array:
"""The log Jacobian of the constraining transforms at `unconstrained`, the packed unconstrained parameters (length `dimension(prepared)`), for the data `prepare_data` returned: `f(u, **prepared)`."""
return model("log_jacobian", unconstrained, **prepared)
def constrain(unconstrained: ArrayLike, **prepared: ArrayLike) -> dict:
"""The constrained value of each parameter, and of each requested value, at `unconstrained`, the packed unconstrained parameters (length `dimension(prepared)`), for the data `prepare_data` returned: `f(u, **prepared)`. Returns a dict, a submodel's values nested under its name:
- `a_Intercept`
- `b_Intercept`
- `b_s_site.tau`
- `b_s_site.L`
- `b_s_site.z`
"""
return model("constrain", unconstrained, **prepared)
def dimension(prepared: dict) -> int:
"""The length of `unconstrained` for the data `prepared` (what `prepare_data` returned)."""
return int(prepared["ppl_last_b_s_site_z"])
def parameters(prepared: dict) -> dict:
"""Each parameter's 0-based `start`/`stop` in `unconstrained`, its transform and dimensions,
for the data `prepared` (what `prepare_data` returned)."""
return {"a_Intercept": {"start": 0, "stop": 1, "transform": "identity", "dims": ()}, "b_Intercept": {"start": 1, "stop": 2, "transform": "identity", "dims": ()}, "b_s_site.tau": {"start": 2, "stop": 4, "transform": "exp", "dims": (2,)}, "b_s_site.L": {"start": 4, "stop": 5, "transform": "lkj", "dims": (2, 2)}, "b_s_site.z": {"start": 5, "stop": int(prepared["ppl_last_b_s_site_z"]), "transform": "identity", "dims": (int(prepared["ppl_axis1_b_s_site_z"]), 2)}}The fits below use 400 sites with 3200 counts in all. The posterior has 805 coordinates: the two population intercepts, the two random-effect sds, the correlation, and two standardized effects per site.
Fit the model on minibatches
brm_svi_fit takes the model on the ReactiveKernels backend, RKBRMI(brmi), with Enzyme loaded for the gradients. Its units are those identified by brm_nlme_view: here one site with its visits. In this model, every observation belongs to exactly one unit, which is what makes a batch of units an unbiased estimate of the whole likelihood.
The approximation has two parts. The coordinates every unit reads, here the population intercepts, the two sds and the correlation, are the globals; their factor is a diagonal Gaussian (family=MeanFieldGlobal(), the default) or a diagonal plus a low-rank part (LowRankGlobal(r)). Each unit's own coordinates, here a site's two standardized effects, get a full Gaussian of their own. The family has no dependence between a unit and the globals, or between two units.
One iteration draws the globals once and a batch of 32 sites. With the default unit_curvature=:orthogonal, each selected two-coordinate site is evaluated at two orthogonal antithetic pairs, giving four unit gradients. Adam updates the global factor and a natural-gradient step updates each selected site's factor. Sites not in the batch do not move. The NUTS reference uses rk_logdensity_problem through WarmupHMC's quorum sampler: four chains are requested, with at least three required to finish. Each starts from its own jittered point near the origin of the unconstrained coordinates.
Quorum selection can favour faster chains; the retained-chain diagnostics cannot rule out that selection bias.
These fitting calls run in the test environment and produce the cached results; the documentation build does not sample.
# SVI: one call. The units are the sites; each iteration draws a batch of 32 of them.
fit_svi(rk; seed=1, progress=nothing) =
brm_svi_fit(rk; batch_size=32, rng=Xoshiro(seed), ndraws=2000, progress)
# NUTS: four adaptive chains of WarmupHMC on the same log density, with a quorum
# of three completed chains and at least 4000 draws each after warm-up.
# Run Julia with one thread for these Enzyme gradients. Each chain starts near the origin of the
# unconstrained coordinates (unit random-effect sds, no site effects), jittered per
# chain, with a unit diagonal metric.
function fit_nuts(rk; seed=10, chains=4, n_draws=4000, progress=nothing)
problem = rk_logdensity_problem(rk; ad_backend=AutoEnzyme(; mode=Enzyme.Reverse))
d = LogDensityProblems.dimension(problem)
rngs = [Xoshiro(seed + c) for c in 1:chains]
init = [(; position=0.1 .* randn(rng, d), squared_scale=ones(d)) for rng in rngs]
completion_warmup_mcmc(rngs, problem;
init, n_draws, min_completed=max(1, div(3chains, 4)), progress).results
endBoth return draws in the packed coordinates of the RK model, one column per draw, so the same readers take them. rk_restore_draws(rk, draws) restores the named parameters on their natural scale; the random-effect block comes back as its margin sds, the Cholesky factor of their correlation and the standardized site draws:
# The population parameters of `site_counts_model`, read from restored draws, each as
# a vector over draws on its natural scale. The random-effect block `s` of the two
# predictors restores as `b_s_site`: `tau` (margin sds, 2 × draws), `L` (one lower
# Cholesky factor of the margins' correlation per draw) and `z` (one sites × 2 matrix
# of standardized site draws per draw). Margin 1 is `a`, margin 2 is `b`.
function population(restored)
block = restored.b_s_site
(; a=restored.a_Intercept, b=restored.b_Intercept,
sd_a=block.tau[1, :], sd_b=block.tau[2, :],
cor_ab=[(L * L')[1, 2] for L in block.L])
end
# Each site's intercept and slope deviation from the population values, sites × draws.
function site_effects(restored)
block = restored.b_s_site
E = [block.tau[:, k] .* (block.L[k] * block.z[k]') for k in eachindex(block.L)]
(; a=reduce(hcat, (e[1, :] for e in E)), b=reduce(hcat, (e[2, :] for e in E)))
endHow the fit stops
The optimization runs in windows of about N / B iterations, at least 100, and stops on its own in three phases. Exploring steps the globals at a constant rate until a window's 10 % trimmed ELBO mean has stopped rising, with no upward trend over the last windows. Settling steps by a fraction of each global's own scale, until the same test passes again or its window limit is reached. Polishing decays the step over 20 windows and returns the average of its iterates. diagnostics.converged reports that the rule fired, which is not a proof that a slow drift has settled.
{"spec":"{\"$schema\":\"https://vega.github.io/schema/vega-lite/v5.json\",\"_aov\":{\"legendInteraction\":\"highlight\"},\"config\":{\"axis\":{\"labelLimit\":0},\"header\":{\"labelLimit\":0},\"legend\":{\"labelLimit\":0},\"style\":{\"aov-ink\":{\"color\":\"#333\"}}},\"data\":{\"values\":[{\"elbo_trimmed_mean\":-9431.77,\"iteration\":100,\"stage\":\"exploring\",\"window\":1},{\"elbo_trimmed_mean\":-8000.24,\"iteration\":200,\"stage\":\"exploring\",\"window\":2},{\"elbo_trimmed_mean\":-7935.92,\"iteration\":300,\"stage\":\"exploring\",\"window\":3},{\"elbo_trimmed_mean\":-7796.54,\"iteration\":400,\"stage\":\"exploring\",\"window\":4},{\"elbo_trimmed_mean\":-7791.34,\"iteration\":500,\"stage\":\"exploring\",\"window\":5},{\"elbo_trimmed_mean\":-7748.71,\"iteration\":600,\"stage\":\"exploring\",\"window\":6},{\"elbo_trimmed_mean\":-7751.38,\"iteration\":700,\"stage\":\"exploring\",\"window\":7},{\"elbo_trimmed_mean\":-7693.56,\"iteration\":800,\"stage\":\"exploring\",\"window\":8},{\"elbo_trimmed_mean\":-7727.48,\"iteration\":900,\"stage\":\"exploring\",\"window\":9},{\"elbo_trimmed_mean\":-7733.9,\"iteration\":1000,\"stage\":\"exploring\",\"window\":10},{\"elbo_trimmed_mean\":-7694.93,\"iteration\":1100,\"stage\":\"exploring\",\"window\":11},{\"elbo_trimmed_mean\":-7717.91,\"iteration\":1200,\"stage\":\"exploring\",\"window\":12},{\"elbo_trimmed_mean\":-7742.7,\"iteration\":1300,\"stage\":\"exploring\",\"window\":13},{\"elbo_trimmed_mean\":-7712.74,\"iteration\":1400,\"stage\":\"exploring\",\"window\":14},{\"elbo_trimmed_mean\":-7702.74,\"iteration\":1500,\"stage\":\"exploring\",\"window\":15},{\"elbo_trimmed_mean\":-7706.99,\"iteration\":1600,\"stage\":\"exploring\",\"window\":16},{\"elbo_trimmed_mean\":-7708.87,\"iteration\":1700,\"stage\":\"exploring\",\"window\":17},{\"elbo_trimmed_mean\":-7724.44,\"iteration\":1800,\"stage\":\"exploring\",\"window\":18},{\"elbo_trimmed_mean\":-7700.17,\"iteration\":1900,\"stage\":\"exploring\",\"window\":19},{\"elbo_trimmed_mean\":-7694.98,\"iteration\":2000,\"stage\":\"exploring\",\"window\":20},{\"elbo_trimmed_mean\":-7704.05,\"iteration\":2100,\"stage\":\"exploring\",\"window\":21},{\"elbo_trimmed_mean\":-7718.59,\"iteration\":2200,\"stage\":\"exploring\",\"window\":22},{\"elbo_trimmed_mean\":-7715.33,\"iteration\":2300,\"stage\":\"exploring\",\"window\":23},{\"elbo_trimmed_mean\":-7677.07,\"iteration\":2400,\"stage\":\"exploring\",\"window\":24},{\"elbo_trimmed_mean\":-7704.71,\"iteration\":2500,\"stage\":\"exploring\",\"window\":25},{\"elbo_trimmed_mean\":-7705.58,\"iteration\":2600,\"stage\":\"exploring\",\"window\":26},{\"elbo_trimmed_mean\":-7674.49,\"iteration\":2700,\"stage\":\"exploring\",\"window\":27},{\"elbo_trimmed_mean\":-7634.51,\"iteration\":2800,\"stage\":\"exploring\",\"window\":28},{\"elbo_trimmed_mean\":-7703.04,\"iteration\":2900,\"stage\":\"exploring\",\"window\":29},{\"elbo_trimmed_mean\":-7710.58,\"iteration\":3000,\"stage\":\"exploring\",\"window\":30},{\"elbo_trimmed_mean\":-7725.01,\"iteration\":3100,\"stage\":\"exploring\",\"window\":31},{\"elbo_trimmed_mean\":-7713.36,\"iteration\":3200,\"stage\":\"exploring\",\"window\":32},{\"elbo_trimmed_mean\":-7720.45,\"iteration\":3300,\"stage\":\"exploring\",\"window\":33},{\"elbo_trimmed_mean\":-7684.59,\"iteration\":3400,\"stage\":\"exploring\",\"window\":34},{\"elbo_trimmed_mean\":-7719.25,\"iteration\":3500,\"stage\":\"exploring\",\"window\":35},{\"elbo_trimmed_mean\":-7715.87,\"iteration\":3600,\"stage\":\"exploring\",\"window\":36},{\"elbo_trimmed_mean\":-7708.83,\"iteration\":3700,\"stage\":\"exploring\",\"window\":37},{\"elbo_trimmed_mean\":-7703.33,\"iteration\":3800,\"stage\":\"exploring\",\"window\":38},{\"elbo_trimmed_mean\":-7695.86,\"iteration\":3900,\"stage\":\"exploring\",\"window\":39},{\"elbo_trimmed_mean\":-7690.46,\"iteration\":4000,\"stage\":\"exploring\",\"window\":40},{\"elbo_trimmed_mean\":-7714.08,\"iteration\":4100,\"stage\":\"exploring\",\"window\":41},{\"elbo_trimmed_mean\":-7695.97,\"iteration\":4200,\"stage\":\"exploring\",\"window\":42},{\"elbo_trimmed_mean\":-7671.27,\"iteration\":4300,\"stage\":\"exploring\",\"window\":43},{\"elbo_trimmed_mean\":-7724.25,\"iteration\":4400,\"stage\":\"exploring\",\"window\":44},{\"elbo_trimmed_mean\":-7708.62,\"iteration\":4500,\"stage\":\"exploring\",\"window\":45},{\"elbo_trimmed_mean\":-7729.77,\"iteration\":4600,\"stage\":\"exploring\",\"window\":46},{\"elbo_trimmed_mean\":-7719.92,\"iteration\":4700,\"stage\":\"settling\",\"window\":47},{\"elbo_trimmed_mean\":-7731.93,\"iteration\":4800,\"stage\":\"settling\",\"window\":48},{\"elbo_trimmed_mean\":-7719.49,\"iteration\":4900,\"stage\":\"settling\",\"window\":49},{\"elbo_trimmed_mean\":-7730.44,\"iteration\":5000,\"stage\":\"settling\",\"window\":50},{\"elbo_trimmed_mean\":-7710.23,\"iteration\":5100,\"stage\":\"settling\",\"window\":51},{\"elbo_trimmed_mean\":-7686.84,\"iteration\":5200,\"stage\":\"settling\",\"window\":52},{\"elbo_trimmed_mean\":-7685.04,\"iteration\":5300,\"stage\":\"settling\",\"window\":53},{\"elbo_trimmed_mean\":-7728.62,\"iteration\":5400,\"stage\":\"settling\",\"window\":54},{\"elbo_trimmed_mean\":-7738.99,\"iteration\":5500,\"stage\":\"settling\",\"window\":55},{\"elbo_trimmed_mean\":-7699.41,\"iteration\":5600,\"stage\":\"settling\",\"window\":56},{\"elbo_trimmed_mean\":-7681.67,\"iteration\":5700,\"stage\":\"settling\",\"window\":57}]},\"height\":220,\"layer\":[{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[\"legend_selection\"],\"transform\":[]},\"encoding\":{\"color\":{\"field\":\"stage\",\"scale\":{\"domain\":[\"exploring\",\"settling\"],\"range\":[\"#2a78d6\",\"#eb6834\"]},\"sort\":[\"exploring\",\"settling\"],\"type\":\"nominal\"},\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"stage\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"iteration\",\"type\":\"quantitative\"},{\"field\":\"elbo_trimmed_mean\",\"type\":\"quantitative\"},{\"field\":\"stage\",\"type\":\"nominal\"}],\"x\":{\"field\":\"iteration\",\"scale\":{\"zero\":false},\"type\":\"quantitative\"},\"y\":{\"field\":\"elbo_trimmed_mean\",\"scale\":{\"zero\":false},\"title\":\"ELBO (10% trimmed window mean)\",\"type\":\"quantitative\"}},\"mark\":{\"filled\":true,\"size\":18.0,\"type\":\"point\"},\"params\":[{\"bind\":\"scales\",\"name\":\"grid\",\"select\":{\"encodings\":[\"x\",\"y\"],\"type\":\"interval\"}},{\"bind\":\"legend\",\"name\":\"legend_selection\",\"select\":{\"fields\":[\"stage\"],\"type\":\"point\"}}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[]},\"data\":{\"values\":[{\"elbo_trimmed_mean\":-7727.74,\"iteration\":5800,\"stage\":\"polishing\",\"window\":58},{\"elbo_trimmed_mean\":-7682.85,\"iteration\":5900,\"stage\":\"polishing\",\"window\":59},{\"elbo_trimmed_mean\":-7690.86,\"iteration\":6000,\"stage\":\"polishing\",\"window\":60},{\"elbo_trimmed_mean\":-7723.17,\"iteration\":6100,\"stage\":\"polishing\",\"window\":61},{\"elbo_trimmed_mean\":-7724.95,\"iteration\":6200,\"stage\":\"polishing\",\"window\":62},{\"elbo_trimmed_mean\":-7715.32,\"iteration\":6300,\"stage\":\"polishing\",\"window\":63},{\"elbo_trimmed_mean\":-7731.08,\"iteration\":6400,\"stage\":\"polishing\",\"window\":64},{\"elbo_trimmed_mean\":-7728.85,\"iteration\":6500,\"stage\":\"polishing\",\"window\":65},{\"elbo_trimmed_mean\":-7691.66,\"iteration\":6600,\"stage\":\"polishing\",\"window\":66},{\"elbo_trimmed_mean\":-7722.94,\"iteration\":6700,\"stage\":\"polishing\",\"window\":67},{\"elbo_trimmed_mean\":-7713.19,\"iteration\":6800,\"stage\":\"polishing\",\"window\":68},{\"elbo_trimmed_mean\":-7690.21,\"iteration\":6900,\"stage\":\"polishing\",\"window\":69},{\"elbo_trimmed_mean\":-7654.34,\"iteration\":7000,\"stage\":\"polishing\",\"window\":70},{\"elbo_trimmed_mean\":-7695.09,\"iteration\":7100,\"stage\":\"polishing\",\"window\":71},{\"elbo_trimmed_mean\":-7707.08,\"iteration\":7200,\"stage\":\"polishing\",\"window\":72},{\"elbo_trimmed_mean\":-7707.09,\"iteration\":7300,\"stage\":\"polishing\",\"window\":73},{\"elbo_trimmed_mean\":-7716.95,\"iteration\":7400,\"stage\":\"polishing\",\"window\":74},{\"elbo_trimmed_mean\":-7679.33,\"iteration\":7500,\"stage\":\"polishing\",\"window\":75},{\"elbo_trimmed_mean\":-7695.24,\"iteration\":7600,\"stage\":\"polishing\",\"window\":76},{\"elbo_trimmed_mean\":-7702.0,\"iteration\":7700,\"stage\":\"polishing\",\"window\":77}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"stage\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"iteration\",\"type\":\"quantitative\"},{\"field\":\"elbo_trimmed_mean\",\"type\":\"quantitative\"}],\"x\":{\"field\":\"iteration\",\"scale\":{\"zero\":false},\"type\":\"quantitative\"},\"y\":{\"field\":\"elbo_trimmed_mean\",\"scale\":{\"zero\":false},\"title\":\"ELBO (10% trimmed window mean)\",\"type\":\"quantitative\"}},\"mark\":{\"filled\":true,\"size\":18.0,\"style\":\"aov-ink\",\"type\":\"point\"}},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[]},\"data\":{\"values\":[{\"elbo_trimmed_mean\":-9431.77,\"iteration\":100,\"stage\":\"exploring\",\"window\":1},{\"elbo_trimmed_mean\":-8000.24,\"iteration\":200,\"stage\":\"exploring\",\"window\":2},{\"elbo_trimmed_mean\":-7935.92,\"iteration\":300,\"stage\":\"exploring\",\"window\":3},{\"elbo_trimmed_mean\":-7796.54,\"iteration\":400,\"stage\":\"exploring\",\"window\":4},{\"elbo_trimmed_mean\":-7791.34,\"iteration\":500,\"stage\":\"exploring\",\"window\":5},{\"elbo_trimmed_mean\":-7748.71,\"iteration\":600,\"stage\":\"exploring\",\"window\":6},{\"elbo_trimmed_mean\":-7751.38,\"iteration\":700,\"stage\":\"exploring\",\"window\":7},{\"elbo_trimmed_mean\":-7693.56,\"iteration\":800,\"stage\":\"exploring\",\"window\":8},{\"elbo_trimmed_mean\":-7727.48,\"iteration\":900,\"stage\":\"exploring\",\"window\":9},{\"elbo_trimmed_mean\":-7733.9,\"iteration\":1000,\"stage\":\"exploring\",\"window\":10},{\"elbo_trimmed_mean\":-7694.93,\"iteration\":1100,\"stage\":\"exploring\",\"window\":11},{\"elbo_trimmed_mean\":-7717.91,\"iteration\":1200,\"stage\":\"exploring\",\"window\":12},{\"elbo_trimmed_mean\":-7742.7,\"iteration\":1300,\"stage\":\"exploring\",\"window\":13},{\"elbo_trimmed_mean\":-7712.74,\"iteration\":1400,\"stage\":\"exploring\",\"window\":14},{\"elbo_trimmed_mean\":-7702.74,\"iteration\":1500,\"stage\":\"exploring\",\"window\":15},{\"elbo_trimmed_mean\":-7706.99,\"iteration\":1600,\"stage\":\"exploring\",\"window\":16},{\"elbo_trimmed_mean\":-7708.87,\"iteration\":1700,\"stage\":\"exploring\",\"window\":17},{\"elbo_trimmed_mean\":-7724.44,\"iteration\":1800,\"stage\":\"exploring\",\"window\":18},{\"elbo_trimmed_mean\":-7700.17,\"iteration\":1900,\"stage\":\"exploring\",\"window\":19},{\"elbo_trimmed_mean\":-7694.98,\"iteration\":2000,\"stage\":\"exploring\",\"window\":20},{\"elbo_trimmed_mean\":-7704.05,\"iteration\":2100,\"stage\":\"exploring\",\"window\":21},{\"elbo_trimmed_mean\":-7718.59,\"iteration\":2200,\"stage\":\"exploring\",\"window\":22},{\"elbo_trimmed_mean\":-7715.33,\"iteration\":2300,\"stage\":\"exploring\",\"window\":23},{\"elbo_trimmed_mean\":-7677.07,\"iteration\":2400,\"stage\":\"exploring\",\"window\":24},{\"elbo_trimmed_mean\":-7704.71,\"iteration\":2500,\"stage\":\"exploring\",\"window\":25},{\"elbo_trimmed_mean\":-7705.58,\"iteration\":2600,\"stage\":\"exploring\",\"window\":26},{\"elbo_trimmed_mean\":-7674.49,\"iteration\":2700,\"stage\":\"exploring\",\"window\":27},{\"elbo_trimmed_mean\":-7634.51,\"iteration\":2800,\"stage\":\"exploring\",\"window\":28},{\"elbo_trimmed_mean\":-7703.04,\"iteration\":2900,\"stage\":\"exploring\",\"window\":29},{\"elbo_trimmed_mean\":-7710.58,\"iteration\":3000,\"stage\":\"exploring\",\"window\":30},{\"elbo_trimmed_mean\":-7725.01,\"iteration\":3100,\"stage\":\"exploring\",\"window\":31},{\"elbo_trimmed_mean\":-7713.36,\"iteration\":3200,\"stage\":\"exploring\",\"window\":32},{\"elbo_trimmed_mean\":-7720.45,\"iteration\":3300,\"stage\":\"exploring\",\"window\":33},{\"elbo_trimmed_mean\":-7684.59,\"iteration\":3400,\"stage\":\"exploring\",\"window\":34},{\"elbo_trimmed_mean\":-7719.25,\"iteration\":3500,\"stage\":\"exploring\",\"window\":35},{\"elbo_trimmed_mean\":-7715.87,\"iteration\":3600,\"stage\":\"exploring\",\"window\":36},{\"elbo_trimmed_mean\":-7708.83,\"iteration\":3700,\"stage\":\"exploring\",\"window\":37},{\"elbo_trimmed_mean\":-7703.33,\"iteration\":3800,\"stage\":\"exploring\",\"window\":38},{\"elbo_trimmed_mean\":-7695.86,\"iteration\":3900,\"stage\":\"exploring\",\"window\":39},{\"elbo_trimmed_mean\":-7690.46,\"iteration\":4000,\"stage\":\"exploring\",\"window\":40},{\"elbo_trimmed_mean\":-7714.08,\"iteration\":4100,\"stage\":\"exploring\",\"window\":41},{\"elbo_trimmed_mean\":-7695.97,\"iteration\":4200,\"stage\":\"exploring\",\"window\":42},{\"elbo_trimmed_mean\":-7671.27,\"iteration\":4300,\"stage\":\"exploring\",\"window\":43},{\"elbo_trimmed_mean\":-7724.25,\"iteration\":4400,\"stage\":\"exploring\",\"window\":44},{\"elbo_trimmed_mean\":-7708.62,\"iteration\":4500,\"stage\":\"exploring\",\"window\":45},{\"elbo_trimmed_mean\":-7729.77,\"iteration\":4600,\"stage\":\"exploring\",\"window\":46},{\"elbo_trimmed_mean\":-7719.92,\"iteration\":4700,\"stage\":\"settling\",\"window\":47},{\"elbo_trimmed_mean\":-7731.93,\"iteration\":4800,\"stage\":\"settling\",\"window\":48},{\"elbo_trimmed_mean\":-7719.49,\"iteration\":4900,\"stage\":\"settling\",\"window\":49},{\"elbo_trimmed_mean\":-7730.44,\"iteration\":5000,\"stage\":\"settling\",\"window\":50},{\"elbo_trimmed_mean\":-7710.23,\"iteration\":5100,\"stage\":\"settling\",\"window\":51},{\"elbo_trimmed_mean\":-7686.84,\"iteration\":5200,\"stage\":\"settling\",\"window\":52},{\"elbo_trimmed_mean\":-7685.04,\"iteration\":5300,\"stage\":\"settling\",\"window\":53},{\"elbo_trimmed_mean\":-7728.62,\"iteration\":5400,\"stage\":\"settling\",\"window\":54},{\"elbo_trimmed_mean\":-7738.99,\"iteration\":5500,\"stage\":\"settling\",\"window\":55},{\"elbo_trimmed_mean\":-7699.41,\"iteration\":5600,\"stage\":\"settling\",\"window\":56},{\"elbo_trimmed_mean\":-7681.67,\"iteration\":5700,\"stage\":\"settling\",\"window\":57},{\"elbo_trimmed_mean\":-7727.74,\"iteration\":5800,\"stage\":\"polishing\",\"window\":58},{\"elbo_trimmed_mean\":-7682.85,\"iteration\":5900,\"stage\":\"polishing\",\"window\":59},{\"elbo_trimmed_mean\":-7690.86,\"iteration\":6000,\"stage\":\"polishing\",\"window\":60},{\"elbo_trimmed_mean\":-7723.17,\"iteration\":6100,\"stage\":\"polishing\",\"window\":61},{\"elbo_trimmed_mean\":-7724.95,\"iteration\":6200,\"stage\":\"polishing\",\"window\":62},{\"elbo_trimmed_mean\":-7715.32,\"iteration\":6300,\"stage\":\"polishing\",\"window\":63},{\"elbo_trimmed_mean\":-7731.08,\"iteration\":6400,\"stage\":\"polishing\",\"window\":64},{\"elbo_trimmed_mean\":-7728.85,\"iteration\":6500,\"stage\":\"polishing\",\"window\":65},{\"elbo_trimmed_mean\":-7691.66,\"iteration\":6600,\"stage\":\"polishing\",\"window\":66},{\"elbo_trimmed_mean\":-7722.94,\"iteration\":6700,\"stage\":\"polishing\",\"window\":67},{\"elbo_trimmed_mean\":-7713.19,\"iteration\":6800,\"stage\":\"polishing\",\"window\":68},{\"elbo_trimmed_mean\":-7690.21,\"iteration\":6900,\"stage\":\"polishing\",\"window\":69},{\"elbo_trimmed_mean\":-7654.34,\"iteration\":7000,\"stage\":\"polishing\",\"window\":70},{\"elbo_trimmed_mean\":-7695.09,\"iteration\":7100,\"stage\":\"polishing\",\"window\":71},{\"elbo_trimmed_mean\":-7707.08,\"iteration\":7200,\"stage\":\"polishing\",\"window\":72},{\"elbo_trimmed_mean\":-7707.09,\"iteration\":7300,\"stage\":\"polishing\",\"window\":73},{\"elbo_trimmed_mean\":-7716.95,\"iteration\":7400,\"stage\":\"polishing\",\"window\":74},{\"elbo_trimmed_mean\":-7679.33,\"iteration\":7500,\"stage\":\"polishing\",\"window\":75},{\"elbo_trimmed_mean\":-7695.24,\"iteration\":7600,\"stage\":\"polishing\",\"window\":76},{\"elbo_trimmed_mean\":-7702.0,\"iteration\":7700,\"stage\":\"polishing\",\"window\":77}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"stage\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":0.4},\"value\":0.06},\"tooltip\":[{\"field\":\"iteration\",\"type\":\"quantitative\"},{\"field\":\"elbo_trimmed_mean\",\"type\":\"quantitative\"}],\"x\":{\"field\":\"iteration\",\"scale\":{\"zero\":false},\"type\":\"quantitative\"},\"y\":{\"field\":\"elbo_trimmed_mean\",\"scale\":{\"zero\":false},\"title\":\"ELBO (10% trimmed window mean)\",\"type\":\"quantitative\"}},\"mark\":{\"opacity\":0.4,\"style\":\"aov-ink\",\"type\":\"line\"}}],\"title\":\"Optimization windows\",\"width\":540,\"description\":\"Trimmed ELBO means over optimization windows: blue exploring, orange settling, neutral ink polishing.\"}","caption":"Trimmed ELBO means over optimization windows: blue exploring, orange settling, neutral ink polishing.","alt":"Trimmed ELBO means over optimization windows: blue exploring, orange settling, neutral ink polishing."}The fit ran 7700 iterations in 46 windows exploring, 11 windows settling and 20 windows polishing, evaluating 985600 site gradients, in 3.2 min. The 3 retained NUTS chains took 3.1 min for 12000 draws (249408 gradient evaluations after warm-up): maximum R-hat 1.003, minimum bulk ESS 1558, minimum tail ESS 3471, 0 divergent transitions (0 % of retained draws). Both ran on one CPU of a shared host; the times are what that run took, not a benchmark.
Compare with NUTS
{"spec":"{\"$schema\":\"https://vega.github.io/schema/vega-lite/v5.json\",\"_aov\":{\"legendInteraction\":\"highlight\"},\"_aovSummary\":{\"bands\":[[\"q025\",\"q975\",\"q025/q975\"],[\"q25\",\"q75\",\"q25/q75\"]],\"point_col\":\"median\",\"value_label\":\"median\"},\"columns\":2,\"config\":{\"axis\":{\"labelLimit\":0},\"header\":{\"labelLimit\":0},\"legend\":{\"labelLimit\":0},\"style\":{\"aov-ink\":{\"color\":\"#333\"}}},\"data\":{\"values\":[{\"__src\":\"_s1\",\"generating\":1.0,\"mean\":1.02437,\"median\":1.02465,\"method\":\"NUTS\",\"parameter\":\"intercept a\",\"q025\":0.969103,\"q25\":1.00622,\"q75\":1.04275,\"q975\":1.07748,\"sd\":0.0276032},{\"__src\":\"_s1\",\"generating\":1.0,\"mean\":1.0225,\"median\":1.02273,\"method\":\"SVI\",\"parameter\":\"intercept a\",\"q025\":1.00425,\"q25\":1.01659,\"q75\":1.02854,\"q975\":1.04018,\"sd\":0.0089684},{\"__src\":\"_s1\",\"generating\":0.4,\"mean\":0.42881,\"median\":0.428688,\"method\":\"NUTS\",\"parameter\":\"slope b\",\"q025\":0.392013,\"q25\":0.416622,\"q75\":0.441179,\"q975\":0.464429,\"sd\":0.0183926},{\"__src\":\"_s1\",\"generating\":0.4,\"mean\":0.428238,\"median\":0.428177,\"method\":\"SVI\",\"parameter\":\"slope b\",\"q025\":0.411602,\"q25\":0.422728,\"q75\":0.4341,\"q975\":0.444679,\"sd\":0.00841773},{\"__src\":\"_s1\",\"generating\":0.5,\"mean\":0.501072,\"median\":0.500739,\"method\":\"NUTS\",\"parameter\":\"sd of site intercepts\",\"q025\":0.459998,\"q25\":0.48618,\"q75\":0.515646,\"q975\":0.544706,\"sd\":0.0216425},{\"__src\":\"_s1\",\"generating\":0.5,\"mean\":0.493347,\"median\":0.493305,\"method\":\"SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.476808,\"q25\":0.487689,\"q75\":0.499073,\"q975\":0.509282,\"sd\":0.00834431},{\"__src\":\"_s1\",\"generating\":0.25,\"mean\":0.263479,\"median\":0.262986,\"method\":\"NUTS\",\"parameter\":\"sd of site slopes\",\"q025\":0.233135,\"q25\":0.25266,\"q75\":0.274042,\"q975\":0.295896,\"sd\":0.0159224},{\"__src\":\"_s1\",\"generating\":0.25,\"mean\":0.260906,\"median\":0.260925,\"method\":\"SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.247947,\"q25\":0.256648,\"q75\":0.265214,\"q975\":0.273312,\"sd\":0.00645138},{\"__src\":\"_s1\",\"generating\":0.3,\"mean\":0.318642,\"median\":0.320387,\"method\":\"NUTS\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.175121,\"q25\":0.269769,\"q75\":0.368465,\"q975\":0.455634,\"sd\":0.0721122},{\"__src\":\"_s1\",\"generating\":0.3,\"mean\":0.313807,\"median\":0.314522,\"method\":\"SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.255532,\"q25\":0.295936,\"q75\":0.332585,\"q975\":0.367555,\"sd\":0.0281799},{\"__src\":\"_s2\",\"generating\":1.0,\"parameter\":\"intercept a\"},{\"__src\":\"_s2\",\"generating\":0.4,\"parameter\":\"slope b\"},{\"__src\":\"_s2\",\"generating\":0.5,\"parameter\":\"sd of site intercepts\"},{\"__src\":\"_s2\",\"generating\":0.25,\"parameter\":\"sd of site slopes\"},{\"__src\":\"_s2\",\"generating\":0.3,\"parameter\":\"correlation of intercepts and slopes\"}]},\"facet\":{\"field\":\"parameter\",\"sort\":[\"intercept a\",\"slope b\",\"sd of site intercepts\",\"sd of site slopes\",\"correlation of intercepts and slopes\"],\"type\":\"nominal\"},\"resolve\":{\"scale\":{\"x\":\"independent\"}},\"spec\":{\"height\":100,\"layer\":[{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[\"legend_selection\"],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},\"encoding\":{\"color\":{\"field\":\"method\",\"scale\":{\"domain\":[\"NUTS\",\"SVI\"],\"range\":[\"#2a78d6\",\"#eb6834\"]},\"type\":\"nominal\"},\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"method\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"field\":\"q025\",\"scale\":{\"zero\":false},\"title\":\"value (dimensionless)\",\"type\":\"quantitative\"},\"x2\":{\"field\":\"q975\"},\"y\":{\"axis\":{\"title\":null},\"field\":\"method\",\"title\":\"\",\"type\":\"nominal\"}},\"mark\":{\"strokeWidth\":1.5,\"type\":\"rule\"},\"params\":[{\"bind\":\"scales\",\"name\":\"grid\",\"select\":{\"encodings\":[\"x\"],\"type\":\"interval\"}},{\"bind\":\"legend\",\"name\":\"legend_selection\",\"select\":{\"fields\":[\"method\"],\"type\":\"point\"}}],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},\"encoding\":{\"color\":{\"field\":\"method\",\"scale\":{\"domain\":[\"NUTS\",\"SVI\"],\"range\":[\"#2a78d6\",\"#eb6834\"]},\"type\":\"nominal\"},\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"method\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"field\":\"q25\",\"scale\":{\"zero\":false},\"title\":\"value (dimensionless)\",\"type\":\"quantitative\"},\"x2\":{\"field\":\"q75\"},\"y\":{\"axis\":{\"title\":null},\"field\":\"method\",\"title\":\"\",\"type\":\"nominal\"}},\"mark\":{\"strokeWidth\":8.0,\"type\":\"rule\"},\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"method\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"field\":\"median\",\"scale\":{\"zero\":false},\"type\":\"quantitative\"},\"y\":{\"field\":\"method\",\"type\":\"nominal\"}},\"mark\":{\"color\":\"white\",\"filled\":true,\"size\":80,\"stroke\":\"#333\",\"strokeWidth\":1.5,\"type\":\"point\"},\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"generating\",\"type\":\"quantitative\"},{\"field\":\"parameter\",\"type\":\"nominal\"}],\"x\":{\"field\":\"generating\",\"scale\":{\"zero\":false},\"type\":\"quantitative\"}},\"mark\":{\"strokeDash\":[6,6],\"style\":\"aov-ink\",\"type\":\"rule\"},\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]}],\"width\":240},\"title\":\"Population parameters\",\"description\":\"Population medians with 50 % and 95 % intervals. Dashed lines mark generating values; SVI intervals are narrower than NUTS intervals.\"}","caption":"Population medians with 50 % and 95 % intervals. Dashed lines mark generating values; SVI intervals are narrower than NUTS intervals.","alt":"Population medians with 50 % and 95 % intervals. Dashed lines mark generating values; SVI intervals are narrower than NUTS intervals."}| parameter | generating value | NUTS: median [95 % interval] | SVI: median [95 % interval] | sd ratio SVI / NUTS |
|---|---|---|---|---|
| intercept a | 1 | 1.02 [0.969, 1.08] | 1.02 [1, 1.04] | 0.32 |
| slope b | 0.4 | 0.429 [0.392, 0.464] | 0.428 [0.412, 0.445] | 0.46 |
| sd of site intercepts | 0.5 | 0.501 [0.46, 0.545] | 0.493 [0.477, 0.509] | 0.39 |
| sd of site slopes | 0.25 | 0.263 [0.233, 0.296] | 0.261 [0.248, 0.273] | 0.41 |
| correlation of intercepts and slopes | 0.3 | 0.32 [0.175, 0.456] | 0.315 [0.256, 0.368] | 0.39 |
The population medians are close, but every SVI marginal is narrower. The factorized Gaussian omits dependence between the globals and the site effects. Nothing in this fit corrects or reweights its draws. Use full MCMC on a feasible reference dataset to assess population uncertainty.
{"spec":"{\"$schema\":\"https://vega.github.io/schema/vega-lite/v5.json\",\"_aov\":{\"autoColumns\":true},\"columns\":2,\"config\":{\"axis\":{\"labelLimit\":0},\"header\":{\"labelLimit\":0},\"legend\":{\"labelLimit\":0},\"style\":{\"aov-ink\":{\"color\":\"#333\"}}},\"data\":{\"values\":[{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0352916,\"nuts_mean\":-0.171347,\"nuts_sd\":0.206642,\"site\":\"s1\",\"svi_mean\":-0.171946,\"svi_sd\":0.211241},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.403426,\"nuts_mean\":-0.464036,\"nuts_sd\":0.256865,\"site\":\"s2\",\"svi_mean\":-0.458874,\"svi_sd\":0.249375},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.582437,\"nuts_mean\":-0.0634789,\"nuts_sd\":0.195649,\"site\":\"s3\",\"svi_mean\":-0.071226,\"svi_sd\":0.194369},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.874967,\"nuts_mean\":-0.537239,\"nuts_sd\":0.256893,\"site\":\"s4\",\"svi_mean\":-0.534491,\"svi_sd\":0.25297},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.521376,\"nuts_mean\":0.369768,\"nuts_sd\":0.164153,\"site\":\"s5\",\"svi_mean\":0.370362,\"svi_sd\":0.166042},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.241126,\"nuts_mean\":0.763077,\"nuts_sd\":0.15584,\"site\":\"s6\",\"svi_mean\":0.762627,\"svi_sd\":0.149897},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.240346,\"nuts_mean\":-0.313635,\"nuts_sd\":0.216679,\"site\":\"s7\",\"svi_mean\":-0.308935,\"svi_sd\":0.22015},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.717306,\"nuts_mean\":-0.951008,\"nuts_sd\":0.288321,\"site\":\"s8\",\"svi_mean\":-0.937302,\"svi_sd\":0.283761},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.121924,\"nuts_mean\":-0.463157,\"nuts_sd\":0.263616,\"site\":\"s9\",\"svi_mean\":-0.453679,\"svi_sd\":0.25654},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.17673,\"nuts_mean\":-0.0757731,\"nuts_sd\":0.197914,\"site\":\"s10\",\"svi_mean\":-0.0709158,\"svi_sd\":0.199617},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.398965,\"nuts_mean\":0.50348,\"nuts_sd\":0.163573,\"site\":\"s11\",\"svi_mean\":0.500878,\"svi_sd\":0.160941},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.252182,\"nuts_mean\":-0.104089,\"nuts_sd\":0.229756,\"site\":\"s12\",\"svi_mean\":-0.0992779,\"svi_sd\":0.236924},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.253504,\"nuts_mean\":0.0988488,\"nuts_sd\":0.194651,\"site\":\"s13\",\"svi_mean\":0.0919949,\"svi_sd\":0.19103},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.413738,\"nuts_mean\":-0.521517,\"nuts_sd\":0.235437,\"site\":\"s14\",\"svi_mean\":-0.516725,\"svi_sd\":0.228506},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.786466,\"nuts_mean\":-0.00603462,\"nuts_sd\":0.200705,\"site\":\"s15\",\"svi_mean\":-0.0123018,\"svi_sd\":0.198613},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.433681,\"nuts_mean\":-0.683179,\"nuts_sd\":0.284593,\"site\":\"s16\",\"svi_mean\":-0.666228,\"svi_sd\":0.277715},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.486012,\"nuts_mean\":0.518851,\"nuts_sd\":0.15949,\"site\":\"s17\",\"svi_mean\":0.512034,\"svi_sd\":0.159711},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.292099,\"nuts_mean\":-0.568371,\"nuts_sd\":0.281392,\"site\":\"s18\",\"svi_mean\":-0.56647,\"svi_sd\":0.271972},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.19576,\"nuts_mean\":0.351413,\"nuts_sd\":0.165574,\"site\":\"s19\",\"svi_mean\":0.358778,\"svi_sd\":0.160555},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.015658,\"nuts_mean\":0.183952,\"nuts_sd\":0.193821,\"site\":\"s20\",\"svi_mean\":0.191141,\"svi_sd\":0.19648},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.382975,\"nuts_mean\":-0.640562,\"nuts_sd\":0.249158,\"site\":\"s21\",\"svi_mean\":-0.637933,\"svi_sd\":0.247175},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.376567,\"nuts_mean\":-0.48434,\"nuts_sd\":0.221752,\"site\":\"s22\",\"svi_mean\":-0.488598,\"svi_sd\":0.220347},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0745305,\"nuts_mean\":0.0104653,\"nuts_sd\":0.20964,\"site\":\"s23\",\"svi_mean\":0.014075,\"svi_sd\":0.200932},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.538123,\"nuts_mean\":-0.151747,\"nuts_sd\":0.213371,\"site\":\"s24\",\"svi_mean\":-0.147367,\"svi_sd\":0.206272},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.235049,\"nuts_mean\":0.724235,\"nuts_sd\":0.144327,\"site\":\"s25\",\"svi_mean\":0.725318,\"svi_sd\":0.145018},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.081675,\"nuts_mean\":0.112232,\"nuts_sd\":0.227719,\"site\":\"s26\",\"svi_mean\":0.108415,\"svi_sd\":0.228314},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.119007,\"nuts_mean\":0.351157,\"nuts_sd\":0.168658,\"site\":\"s27\",\"svi_mean\":0.35367,\"svi_sd\":0.167955},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.845388,\"nuts_mean\":-0.106215,\"nuts_sd\":0.199654,\"site\":\"s28\",\"svi_mean\":-0.107111,\"svi_sd\":0.201688},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.457741,\"nuts_mean\":-0.272663,\"nuts_sd\":0.210376,\"site\":\"s29\",\"svi_mean\":-0.279865,\"svi_sd\":0.210629},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.37853,\"nuts_mean\":-0.835285,\"nuts_sd\":0.267727,\"site\":\"s30\",\"svi_mean\":-0.822588,\"svi_sd\":0.270495},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.302105,\"nuts_mean\":-0.0582827,\"nuts_sd\":0.19371,\"site\":\"s31\",\"svi_mean\":-0.0610986,\"svi_sd\":0.193263},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.444077,\"nuts_mean\":-0.18596,\"nuts_sd\":0.207601,\"site\":\"s32\",\"svi_mean\":-0.18525,\"svi_sd\":0.205886},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.138513,\"nuts_mean\":-0.451444,\"nuts_sd\":0.228249,\"site\":\"s33\",\"svi_mean\":-0.447975,\"svi_sd\":0.226192},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.995506,\"nuts_mean\":-0.0494383,\"nuts_sd\":0.205046,\"site\":\"s34\",\"svi_mean\":-0.0507243,\"svi_sd\":0.203027},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.376526,\"nuts_mean\":0.119482,\"nuts_sd\":0.249502,\"site\":\"s35\",\"svi_mean\":0.112016,\"svi_sd\":0.245073},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.186485,\"nuts_mean\":0.670599,\"nuts_sd\":0.172183,\"site\":\"s36\",\"svi_mean\":0.669425,\"svi_sd\":0.170146},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.661724,\"nuts_mean\":0.18722,\"nuts_sd\":0.180497,\"site\":\"s37\",\"svi_mean\":0.190598,\"svi_sd\":0.184363},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.869904,\"nuts_mean\":-0.0701655,\"nuts_sd\":0.196606,\"site\":\"s38\",\"svi_mean\":-0.0659741,\"svi_sd\":0.188651},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.288566,\"nuts_mean\":0.81433,\"nuts_sd\":0.190614,\"site\":\"s39\",\"svi_mean\":0.815665,\"svi_sd\":0.185462},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.169697,\"nuts_mean\":0.335166,\"nuts_sd\":0.171133,\"site\":\"s40\",\"svi_mean\":0.34499,\"svi_sd\":0.171771},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.640722,\"nuts_mean\":0.10623,\"nuts_sd\":0.189468,\"site\":\"s41\",\"svi_mean\":0.111147,\"svi_sd\":0.190285},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-1.58418,\"nuts_mean\":-1.10926,\"nuts_sd\":0.285554,\"site\":\"s42\",\"svi_mean\":-1.08558,\"svi_sd\":0.281276},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.305285,\"nuts_mean\":-0.495639,\"nuts_sd\":0.246395,\"site\":\"s43\",\"svi_mean\":-0.494005,\"svi_sd\":0.239558},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.25333,\"nuts_mean\":-0.559794,\"nuts_sd\":0.243201,\"site\":\"s44\",\"svi_mean\":-0.548138,\"svi_sd\":0.235503},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.842337,\"nuts_mean\":-0.334416,\"nuts_sd\":0.220823,\"site\":\"s45\",\"svi_mean\":-0.335695,\"svi_sd\":0.218132},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.952805,\"nuts_mean\":0.403092,\"nuts_sd\":0.188674,\"site\":\"s46\",\"svi_mean\":0.403165,\"svi_sd\":0.179993},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.603313,\"nuts_mean\":0.241887,\"nuts_sd\":0.19006,\"site\":\"s47\",\"svi_mean\":0.241242,\"svi_sd\":0.193278},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.343921,\"nuts_mean\":0.650549,\"nuts_sd\":0.168623,\"site\":\"s48\",\"svi_mean\":0.646757,\"svi_sd\":0.164128},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.225961,\"nuts_mean\":0.12388,\"nuts_sd\":0.187635,\"site\":\"s49\",\"svi_mean\":0.131634,\"svi_sd\":0.182108},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.486845,\"nuts_mean\":0.126596,\"nuts_sd\":0.183549,\"site\":\"s50\",\"svi_mean\":0.126918,\"svi_sd\":0.178841},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.854298,\"nuts_mean\":0.278446,\"nuts_sd\":0.182093,\"site\":\"s51\",\"svi_mean\":0.277714,\"svi_sd\":0.177322},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0250188,\"nuts_mean\":-0.17069,\"nuts_sd\":0.21586,\"site\":\"s52\",\"svi_mean\":-0.169235,\"svi_sd\":0.208667},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.366407,\"nuts_mean\":0.361915,\"nuts_sd\":0.177173,\"site\":\"s53\",\"svi_mean\":0.368811,\"svi_sd\":0.176077},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.655658,\"nuts_mean\":-0.721999,\"nuts_sd\":0.252801,\"site\":\"s54\",\"svi_mean\":-0.724797,\"svi_sd\":0.247723},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.143729,\"nuts_mean\":0.422059,\"nuts_sd\":0.159904,\"site\":\"s55\",\"svi_mean\":0.43224,\"svi_sd\":0.163512},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.225124,\"nuts_mean\":0.192358,\"nuts_sd\":0.182494,\"site\":\"s56\",\"svi_mean\":0.198234,\"svi_sd\":0.180212},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.143281,\"nuts_mean\":-0.511249,\"nuts_sd\":0.24387,\"site\":\"s57\",\"svi_mean\":-0.50425,\"svi_sd\":0.25053},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.806876,\"nuts_mean\":-0.376504,\"nuts_sd\":0.217553,\"site\":\"s58\",\"svi_mean\":-0.382733,\"svi_sd\":0.218129},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.616978,\"nuts_mean\":-0.391804,\"nuts_sd\":0.222239,\"site\":\"s59\",\"svi_mean\":-0.383526,\"svi_sd\":0.214707},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.211113,\"nuts_mean\":0.467431,\"nuts_sd\":0.174357,\"site\":\"s60\",\"svi_mean\":0.47134,\"svi_sd\":0.172286},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.248651,\"nuts_mean\":0.587118,\"nuts_sd\":0.1548,\"site\":\"s61\",\"svi_mean\":0.583569,\"svi_sd\":0.153432},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.708649,\"nuts_mean\":-0.463496,\"nuts_sd\":0.240965,\"site\":\"s62\",\"svi_mean\":-0.449337,\"svi_sd\":0.237166},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.988934,\"nuts_mean\":0.649184,\"nuts_sd\":0.147687,\"site\":\"s63\",\"svi_mean\":0.650234,\"svi_sd\":0.146678},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0616643,\"nuts_mean\":-0.542348,\"nuts_sd\":0.236924,\"site\":\"s64\",\"svi_mean\":-0.53984,\"svi_sd\":0.232767},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.130676,\"nuts_mean\":-0.0411077,\"nuts_sd\":0.195897,\"site\":\"s65\",\"svi_mean\":-0.0396251,\"svi_sd\":0.192862},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.732499,\"nuts_mean\":-0.268315,\"nuts_sd\":0.224306,\"site\":\"s66\",\"svi_mean\":-0.274658,\"svi_sd\":0.229137},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.278277,\"nuts_mean\":0.16503,\"nuts_sd\":0.174797,\"site\":\"s67\",\"svi_mean\":0.168599,\"svi_sd\":0.177904},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.420198,\"nuts_mean\":0.103227,\"nuts_sd\":0.184379,\"site\":\"s68\",\"svi_mean\":0.102189,\"svi_sd\":0.184524},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0272954,\"nuts_mean\":-0.45572,\"nuts_sd\":0.227884,\"site\":\"s69\",\"svi_mean\":-0.456126,\"svi_sd\":0.218635},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.532083,\"nuts_mean\":1.12509,\"nuts_sd\":0.148013,\"site\":\"s70\",\"svi_mean\":1.12228,\"svi_sd\":0.150292},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.03111,\"nuts_mean\":-0.217473,\"nuts_sd\":0.229951,\"site\":\"s71\",\"svi_mean\":-0.218369,\"svi_sd\":0.22449},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.49102,\"nuts_mean\":-0.451832,\"nuts_sd\":0.235493,\"site\":\"s72\",\"svi_mean\":-0.445054,\"svi_sd\":0.234275},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.622704,\"nuts_mean\":-0.375934,\"nuts_sd\":0.21969,\"site\":\"s73\",\"svi_mean\":-0.379956,\"svi_sd\":0.214786},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.278745,\"nuts_mean\":-0.549623,\"nuts_sd\":0.243097,\"site\":\"s74\",\"svi_mean\":-0.540704,\"svi_sd\":0.243235},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.141257,\"nuts_mean\":-0.0632537,\"nuts_sd\":0.194322,\"site\":\"s75\",\"svi_mean\":-0.0677106,\"svi_sd\":0.193158},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.597343,\"nuts_mean\":-0.227667,\"nuts_sd\":0.211141,\"site\":\"s76\",\"svi_mean\":-0.224385,\"svi_sd\":0.213338},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.239716,\"nuts_mean\":0.871281,\"nuts_sd\":0.136046,\"site\":\"s77\",\"svi_mean\":0.869373,\"svi_sd\":0.130132},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.06599,\"nuts_mean\":-0.207176,\"nuts_sd\":0.213915,\"site\":\"s78\",\"svi_mean\":-0.200618,\"svi_sd\":0.208368},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.176621,\"nuts_mean\":0.500289,\"nuts_sd\":0.168251,\"site\":\"s79\",\"svi_mean\":0.500134,\"svi_sd\":0.160823},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0578997,\"nuts_mean\":0.347163,\"nuts_sd\":0.186143,\"site\":\"s80\",\"svi_mean\":0.331982,\"svi_sd\":0.183827},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.116383,\"nuts_mean\":-0.143422,\"nuts_sd\":0.205704,\"site\":\"s81\",\"svi_mean\":-0.145414,\"svi_sd\":0.208441},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.112829,\"nuts_mean\":0.270382,\"nuts_sd\":0.184918,\"site\":\"s82\",\"svi_mean\":0.269128,\"svi_sd\":0.183464},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.027912,\"nuts_mean\":-0.427495,\"nuts_sd\":0.222641,\"site\":\"s83\",\"svi_mean\":-0.420841,\"svi_sd\":0.222159},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.540746,\"nuts_mean\":0.045176,\"nuts_sd\":0.188112,\"site\":\"s84\",\"svi_mean\":0.0424379,\"svi_sd\":0.175949},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.639473,\"nuts_mean\":-0.663059,\"nuts_sd\":0.245928,\"site\":\"s85\",\"svi_mean\":-0.661886,\"svi_sd\":0.240899},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0630453,\"nuts_mean\":-0.756257,\"nuts_sd\":0.258304,\"site\":\"s86\",\"svi_mean\":-0.735342,\"svi_sd\":0.250741},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.598095,\"nuts_mean\":0.139789,\"nuts_sd\":0.179081,\"site\":\"s87\",\"svi_mean\":0.135913,\"svi_sd\":0.173662},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.551454,\"nuts_mean\":-0.47941,\"nuts_sd\":0.252008,\"site\":\"s88\",\"svi_mean\":-0.47129,\"svi_sd\":0.24017},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.317194,\"nuts_mean\":-0.185903,\"nuts_sd\":0.223429,\"site\":\"s89\",\"svi_mean\":-0.18688,\"svi_sd\":0.220357},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.563838,\"nuts_mean\":-0.172188,\"nuts_sd\":0.207791,\"site\":\"s90\",\"svi_mean\":-0.172303,\"svi_sd\":0.203609},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.167297,\"nuts_mean\":-0.212474,\"nuts_sd\":0.204217,\"site\":\"s91\",\"svi_mean\":-0.208714,\"svi_sd\":0.205171},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.427003,\"nuts_mean\":0.630739,\"nuts_sd\":0.179754,\"site\":\"s92\",\"svi_mean\":0.628979,\"svi_sd\":0.183378},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.53629,\"nuts_mean\":-0.0841889,\"nuts_sd\":0.204869,\"site\":\"s93\",\"svi_mean\":-0.0862104,\"svi_sd\":0.201934},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-1.00145,\"nuts_mean\":0.0231318,\"nuts_sd\":0.192874,\"site\":\"s94\",\"svi_mean\":0.0259683,\"svi_sd\":0.184641},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.457455,\"nuts_mean\":-0.539863,\"nuts_sd\":0.270981,\"site\":\"s95\",\"svi_mean\":-0.542899,\"svi_sd\":0.267731},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0516996,\"nuts_mean\":0.821829,\"nuts_sd\":0.14948,\"site\":\"s96\",\"svi_mean\":0.820844,\"svi_sd\":0.146035},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.678276,\"nuts_mean\":0.077637,\"nuts_sd\":0.187344,\"site\":\"s97\",\"svi_mean\":0.0797038,\"svi_sd\":0.18267},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.867319,\"nuts_mean\":0.143583,\"nuts_sd\":0.18237,\"site\":\"s98\",\"svi_mean\":0.147599,\"svi_sd\":0.178316},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.636917,\"nuts_mean\":-0.291777,\"nuts_sd\":0.257938,\"site\":\"s99\",\"svi_mean\":-0.289534,\"svi_sd\":0.258596},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0683469,\"nuts_mean\":0.0277551,\"nuts_sd\":0.188724,\"site\":\"s100\",\"svi_mean\":0.0351447,\"svi_sd\":0.189463},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.487289,\"nuts_mean\":-0.244612,\"nuts_sd\":0.230456,\"site\":\"s101\",\"svi_mean\":-0.249311,\"svi_sd\":0.23253},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.791362,\"nuts_mean\":-0.163079,\"nuts_sd\":0.199817,\"site\":\"s102\",\"svi_mean\":-0.154315,\"svi_sd\":0.196878},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.847238,\"nuts_mean\":-0.521362,\"nuts_sd\":0.233326,\"site\":\"s103\",\"svi_mean\":-0.514375,\"svi_sd\":0.23067},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.390144,\"nuts_mean\":0.742877,\"nuts_sd\":0.167715,\"site\":\"s104\",\"svi_mean\":0.737961,\"svi_sd\":0.161895},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.640901,\"nuts_mean\":-0.258696,\"nuts_sd\":0.216268,\"site\":\"s105\",\"svi_mean\":-0.254765,\"svi_sd\":0.218509},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.702339,\"nuts_mean\":-0.105777,\"nuts_sd\":0.205797,\"site\":\"s106\",\"svi_mean\":-0.0948306,\"svi_sd\":0.204296},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0449937,\"nuts_mean\":0.0844747,\"nuts_sd\":0.19549,\"site\":\"s107\",\"svi_mean\":0.0858392,\"svi_sd\":0.187865},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.446697,\"nuts_mean\":-0.0507441,\"nuts_sd\":0.213734,\"site\":\"s108\",\"svi_mean\":-0.0530042,\"svi_sd\":0.215069},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.356627,\"nuts_mean\":-0.496992,\"nuts_sd\":0.245803,\"site\":\"s109\",\"svi_mean\":-0.496029,\"svi_sd\":0.234868},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.406109,\"nuts_mean\":0.785414,\"nuts_sd\":0.168129,\"site\":\"s110\",\"svi_mean\":0.782431,\"svi_sd\":0.16468},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.117646,\"nuts_mean\":0.139425,\"nuts_sd\":0.183809,\"site\":\"s111\",\"svi_mean\":0.140437,\"svi_sd\":0.183862},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.55574,\"nuts_mean\":-0.0995712,\"nuts_sd\":0.198096,\"site\":\"s112\",\"svi_mean\":-0.101874,\"svi_sd\":0.193191},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.639468,\"nuts_mean\":-0.37415,\"nuts_sd\":0.225303,\"site\":\"s113\",\"svi_mean\":-0.371925,\"svi_sd\":0.221355},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.631705,\"nuts_mean\":-0.00400934,\"nuts_sd\":0.20175,\"site\":\"s114\",\"svi_mean\":-0.00219941,\"svi_sd\":0.206573},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.381867,\"nuts_mean\":0.240657,\"nuts_sd\":0.174919,\"site\":\"s115\",\"svi_mean\":0.239625,\"svi_sd\":0.178753},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.447651,\"nuts_mean\":0.776998,\"nuts_sd\":0.140014,\"site\":\"s116\",\"svi_mean\":0.776989,\"svi_sd\":0.139473},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.792362,\"nuts_mean\":0.154995,\"nuts_sd\":0.183377,\"site\":\"s117\",\"svi_mean\":0.149132,\"svi_sd\":0.181374},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.51296,\"nuts_mean\":0.315163,\"nuts_sd\":0.2303,\"site\":\"s118\",\"svi_mean\":0.316427,\"svi_sd\":0.232849},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0372457,\"nuts_mean\":0.318139,\"nuts_sd\":0.169417,\"site\":\"s119\",\"svi_mean\":0.321531,\"svi_sd\":0.165381},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.745496,\"nuts_mean\":0.380933,\"nuts_sd\":0.163189,\"site\":\"s120\",\"svi_mean\":0.38364,\"svi_sd\":0.159224},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.354587,\"nuts_mean\":0.0932809,\"nuts_sd\":0.213486,\"site\":\"s121\",\"svi_mean\":0.0910309,\"svi_sd\":0.207212},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.298126,\"nuts_mean\":-0.377256,\"nuts_sd\":0.22638,\"site\":\"s122\",\"svi_mean\":-0.380185,\"svi_sd\":0.226832},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0439006,\"nuts_mean\":1.52797,\"nuts_sd\":0.107787,\"site\":\"s123\",\"svi_mean\":1.53177,\"svi_sd\":0.109044},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0956233,\"nuts_mean\":-0.269521,\"nuts_sd\":0.21136,\"site\":\"s124\",\"svi_mean\":-0.27193,\"svi_sd\":0.218038},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.30345,\"nuts_mean\":0.404275,\"nuts_sd\":0.169654,\"site\":\"s125\",\"svi_mean\":0.405824,\"svi_sd\":0.166603},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0195375,\"nuts_mean\":0.781429,\"nuts_sd\":0.139768,\"site\":\"s126\",\"svi_mean\":0.782974,\"svi_sd\":0.134373},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.366783,\"nuts_mean\":0.627757,\"nuts_sd\":0.1513,\"site\":\"s127\",\"svi_mean\":0.628325,\"svi_sd\":0.151816},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.381491,\"nuts_mean\":-0.194425,\"nuts_sd\":0.206551,\"site\":\"s128\",\"svi_mean\":-0.17995,\"svi_sd\":0.203606},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.183262,\"nuts_mean\":0.272312,\"nuts_sd\":0.205338,\"site\":\"s129\",\"svi_mean\":0.272896,\"svi_sd\":0.203494},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.503093,\"nuts_mean\":-0.10921,\"nuts_sd\":0.192561,\"site\":\"s130\",\"svi_mean\":-0.109087,\"svi_sd\":0.19258},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.397923,\"nuts_mean\":-0.385335,\"nuts_sd\":0.223869,\"site\":\"s131\",\"svi_mean\":-0.380793,\"svi_sd\":0.218853},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0169937,\"nuts_mean\":0.669189,\"nuts_sd\":0.148712,\"site\":\"s132\",\"svi_mean\":0.66965,\"svi_sd\":0.147795},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.682243,\"nuts_mean\":0.205441,\"nuts_sd\":0.184053,\"site\":\"s133\",\"svi_mean\":0.207681,\"svi_sd\":0.18007},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.414062,\"nuts_mean\":-0.407016,\"nuts_sd\":0.241081,\"site\":\"s134\",\"svi_mean\":-0.400963,\"svi_sd\":0.237018},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.225953,\"nuts_mean\":-0.618086,\"nuts_sd\":0.247672,\"site\":\"s135\",\"svi_mean\":-0.610448,\"svi_sd\":0.242395},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.919514,\"nuts_mean\":0.0756343,\"nuts_sd\":0.187931,\"site\":\"s136\",\"svi_mean\":0.0793435,\"svi_sd\":0.182925},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.617914,\"nuts_mean\":0.106259,\"nuts_sd\":0.189677,\"site\":\"s137\",\"svi_mean\":0.104251,\"svi_sd\":0.184704},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.685274,\"nuts_mean\":-0.669393,\"nuts_sd\":0.256829,\"site\":\"s138\",\"svi_mean\":-0.65765,\"svi_sd\":0.248775},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.702179,\"nuts_mean\":0.204876,\"nuts_sd\":0.184148,\"site\":\"s139\",\"svi_mean\":0.207805,\"svi_sd\":0.179446},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.231478,\"nuts_mean\":-0.114986,\"nuts_sd\":0.207627,\"site\":\"s140\",\"svi_mean\":-0.114579,\"svi_sd\":0.202546},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.315266,\"nuts_mean\":0.36987,\"nuts_sd\":0.211013,\"site\":\"s141\",\"svi_mean\":0.37718,\"svi_sd\":0.210935},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.131008,\"nuts_mean\":-0.499769,\"nuts_sd\":0.23324,\"site\":\"s142\",\"svi_mean\":-0.491397,\"svi_sd\":0.236041},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.21919,\"nuts_mean\":-0.371393,\"nuts_sd\":0.226401,\"site\":\"s143\",\"svi_mean\":-0.355366,\"svi_sd\":0.224406},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.488598,\"nuts_mean\":0.0625494,\"nuts_sd\":0.191825,\"site\":\"s144\",\"svi_mean\":0.0666911,\"svi_sd\":0.189171},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.465921,\"nuts_mean\":0.262193,\"nuts_sd\":0.170257,\"site\":\"s145\",\"svi_mean\":0.258231,\"svi_sd\":0.168388},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.306116,\"nuts_mean\":-0.32235,\"nuts_sd\":0.220291,\"site\":\"s146\",\"svi_mean\":-0.322029,\"svi_sd\":0.221757},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.344287,\"nuts_mean\":-0.489522,\"nuts_sd\":0.226615,\"site\":\"s147\",\"svi_mean\":-0.480183,\"svi_sd\":0.228151},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.355928,\"nuts_mean\":0.498026,\"nuts_sd\":0.160724,\"site\":\"s148\",\"svi_mean\":0.493912,\"svi_sd\":0.158548},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.199192,\"nuts_mean\":0.45338,\"nuts_sd\":0.193676,\"site\":\"s149\",\"svi_mean\":0.465668,\"svi_sd\":0.189754},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0232092,\"nuts_mean\":0.0882457,\"nuts_sd\":0.185502,\"site\":\"s150\",\"svi_mean\":0.092646,\"svi_sd\":0.183026},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.241924,\"nuts_mean\":0.0567495,\"nuts_sd\":0.210606,\"site\":\"s151\",\"svi_mean\":0.0546007,\"svi_sd\":0.206717},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.645684,\"nuts_mean\":0.124639,\"nuts_sd\":0.183981,\"site\":\"s152\",\"svi_mean\":0.127252,\"svi_sd\":0.183194},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.813277,\"nuts_mean\":-0.29012,\"nuts_sd\":0.224844,\"site\":\"s153\",\"svi_mean\":-0.285817,\"svi_sd\":0.221507},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.546764,\"nuts_mean\":-0.228279,\"nuts_sd\":0.213759,\"site\":\"s154\",\"svi_mean\":-0.23912,\"svi_sd\":0.2026},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.93118,\"nuts_mean\":-0.457081,\"nuts_sd\":0.255327,\"site\":\"s155\",\"svi_mean\":-0.456097,\"svi_sd\":0.256617},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.672008,\"nuts_mean\":-0.706021,\"nuts_sd\":0.25392,\"site\":\"s156\",\"svi_mean\":-0.700417,\"svi_sd\":0.248765},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.267292,\"nuts_mean\":0.394966,\"nuts_sd\":0.207277,\"site\":\"s157\",\"svi_mean\":0.396508,\"svi_sd\":0.200526},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0246877,\"nuts_mean\":-0.576089,\"nuts_sd\":0.239077,\"site\":\"s158\",\"svi_mean\":-0.569953,\"svi_sd\":0.239671},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.115545,\"nuts_mean\":0.88322,\"nuts_sd\":0.139518,\"site\":\"s159\",\"svi_mean\":0.885118,\"svi_sd\":0.134177},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.732768,\"nuts_mean\":-0.015215,\"nuts_sd\":0.22591,\"site\":\"s160\",\"svi_mean\":-0.00939248,\"svi_sd\":0.227296},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.30922,\"nuts_mean\":-0.210036,\"nuts_sd\":0.222151,\"site\":\"s161\",\"svi_mean\":-0.198659,\"svi_sd\":0.215966},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.203309,\"nuts_mean\":-0.238946,\"nuts_sd\":0.213585,\"site\":\"s162\",\"svi_mean\":-0.2268,\"svi_sd\":0.211765},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.420984,\"nuts_mean\":-0.547482,\"nuts_sd\":0.245463,\"site\":\"s163\",\"svi_mean\":-0.547496,\"svi_sd\":0.241803},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.538871,\"nuts_mean\":0.321983,\"nuts_sd\":0.168115,\"site\":\"s164\",\"svi_mean\":0.322822,\"svi_sd\":0.163566},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0457585,\"nuts_mean\":-0.484774,\"nuts_sd\":0.278735,\"site\":\"s165\",\"svi_mean\":-0.476427,\"svi_sd\":0.279808},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.162206,\"nuts_mean\":-0.248582,\"nuts_sd\":0.210801,\"site\":\"s166\",\"svi_mean\":-0.238374,\"svi_sd\":0.217896},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.162032,\"nuts_mean\":0.0625899,\"nuts_sd\":0.206546,\"site\":\"s167\",\"svi_mean\":0.0603617,\"svi_sd\":0.201664},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.856028,\"nuts_mean\":-0.0981746,\"nuts_sd\":0.204193,\"site\":\"s168\",\"svi_mean\":-0.102229,\"svi_sd\":0.20386},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.263993,\"nuts_mean\":-0.140453,\"nuts_sd\":0.201953,\"site\":\"s169\",\"svi_mean\":-0.1366,\"svi_sd\":0.200183},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.665923,\"nuts_mean\":-0.275207,\"nuts_sd\":0.217439,\"site\":\"s170\",\"svi_mean\":-0.26793,\"svi_sd\":0.213802},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.402822,\"nuts_mean\":-0.318795,\"nuts_sd\":0.220951,\"site\":\"s171\",\"svi_mean\":-0.308,\"svi_sd\":0.223554},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.32159,\"nuts_mean\":0.247118,\"nuts_sd\":0.177836,\"site\":\"s172\",\"svi_mean\":0.255013,\"svi_sd\":0.169262},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0540445,\"nuts_mean\":-0.572513,\"nuts_sd\":0.246436,\"site\":\"s173\",\"svi_mean\":-0.575184,\"svi_sd\":0.242588},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0479579,\"nuts_mean\":0.0605418,\"nuts_sd\":0.243151,\"site\":\"s174\",\"svi_mean\":0.0603397,\"svi_sd\":0.246274},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.508629,\"nuts_mean\":0.103721,\"nuts_sd\":0.198839,\"site\":\"s175\",\"svi_mean\":0.116184,\"svi_sd\":0.193044},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.440407,\"nuts_mean\":0.0790476,\"nuts_sd\":0.19495,\"site\":\"s176\",\"svi_mean\":0.0887738,\"svi_sd\":0.190585},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.322876,\"nuts_mean\":0.543168,\"nuts_sd\":0.159594,\"site\":\"s177\",\"svi_mean\":0.542849,\"svi_sd\":0.161279},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0681554,\"nuts_mean\":-0.563013,\"nuts_sd\":0.234295,\"site\":\"s178\",\"svi_mean\":-0.56271,\"svi_sd\":0.226705},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.487763,\"nuts_mean\":-0.548137,\"nuts_sd\":0.238744,\"site\":\"s179\",\"svi_mean\":-0.5328,\"svi_sd\":0.233856},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.131546,\"nuts_mean\":-0.0888363,\"nuts_sd\":0.203351,\"site\":\"s180\",\"svi_mean\":-0.0863961,\"svi_sd\":0.202387},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.23928,\"nuts_mean\":-0.533404,\"nuts_sd\":0.234156,\"site\":\"s181\",\"svi_mean\":-0.531448,\"svi_sd\":0.228972},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.23109,\"nuts_mean\":-0.374285,\"nuts_sd\":0.23087,\"site\":\"s182\",\"svi_mean\":-0.362415,\"svi_sd\":0.230519},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0116997,\"nuts_mean\":-0.0223285,\"nuts_sd\":0.205874,\"site\":\"s183\",\"svi_mean\":-0.0162751,\"svi_sd\":0.210419},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.282896,\"nuts_mean\":0.331491,\"nuts_sd\":0.167658,\"site\":\"s184\",\"svi_mean\":0.329146,\"svi_sd\":0.17041},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.816155,\"nuts_mean\":0.257209,\"nuts_sd\":0.185154,\"site\":\"s185\",\"svi_mean\":0.265086,\"svi_sd\":0.180465},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.288012,\"nuts_mean\":1.29622,\"nuts_sd\":0.119796,\"site\":\"s186\",\"svi_mean\":1.29481,\"svi_sd\":0.116648},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.188254,\"nuts_mean\":-0.543742,\"nuts_sd\":0.252184,\"site\":\"s187\",\"svi_mean\":-0.528937,\"svi_sd\":0.245843},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.200868,\"nuts_mean\":0.183797,\"nuts_sd\":0.184233,\"site\":\"s188\",\"svi_mean\":0.179245,\"svi_sd\":0.185063},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.124084,\"nuts_mean\":0.26932,\"nuts_sd\":0.174383,\"site\":\"s189\",\"svi_mean\":0.274129,\"svi_sd\":0.174344},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.00628113,\"nuts_mean\":-0.735924,\"nuts_sd\":0.2471,\"site\":\"s190\",\"svi_mean\":-0.739726,\"svi_sd\":0.247011},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.889108,\"nuts_mean\":0.0908912,\"nuts_sd\":0.208765,\"site\":\"s191\",\"svi_mean\":0.0890702,\"svi_sd\":0.21205},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.435704,\"nuts_mean\":-0.295328,\"nuts_sd\":0.213359,\"site\":\"s192\",\"svi_mean\":-0.292529,\"svi_sd\":0.212156},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.278669,\"nuts_mean\":0.126899,\"nuts_sd\":0.185448,\"site\":\"s193\",\"svi_mean\":0.128739,\"svi_sd\":0.18167},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.177703,\"nuts_mean\":-0.344189,\"nuts_sd\":0.22075,\"site\":\"s194\",\"svi_mean\":-0.331685,\"svi_sd\":0.22065},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.247916,\"nuts_mean\":-0.510841,\"nuts_sd\":0.239874,\"site\":\"s195\",\"svi_mean\":-0.502818,\"svi_sd\":0.230442},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.48687,\"nuts_mean\":-0.28242,\"nuts_sd\":0.209645,\"site\":\"s196\",\"svi_mean\":-0.276123,\"svi_sd\":0.211112},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.239107,\"nuts_mean\":0.822469,\"nuts_sd\":0.172773,\"site\":\"s197\",\"svi_mean\":0.816179,\"svi_sd\":0.170794},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.977344,\"nuts_mean\":0.0269256,\"nuts_sd\":0.189226,\"site\":\"s198\",\"svi_mean\":0.0324381,\"svi_sd\":0.191849},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.220676,\"nuts_mean\":0.17514,\"nuts_sd\":0.184137,\"site\":\"s199\",\"svi_mean\":0.173182,\"svi_sd\":0.183416},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.164386,\"nuts_mean\":-0.217543,\"nuts_sd\":0.207229,\"site\":\"s200\",\"svi_mean\":-0.210865,\"svi_sd\":0.200045},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.351203,\"nuts_mean\":-0.633475,\"nuts_sd\":0.247321,\"site\":\"s201\",\"svi_mean\":-0.623265,\"svi_sd\":0.250988},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.854284,\"nuts_mean\":0.0244108,\"nuts_sd\":0.20924,\"site\":\"s202\",\"svi_mean\":0.0284411,\"svi_sd\":0.204848},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.336319,\"nuts_mean\":-0.0218097,\"nuts_sd\":0.212174,\"site\":\"s203\",\"svi_mean\":-0.0179771,\"svi_sd\":0.205871},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.195145,\"nuts_mean\":-0.902254,\"nuts_sd\":0.265007,\"site\":\"s204\",\"svi_mean\":-0.89152,\"svi_sd\":0.265996},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.345448,\"nuts_mean\":-0.700257,\"nuts_sd\":0.248962,\"site\":\"s205\",\"svi_mean\":-0.691611,\"svi_sd\":0.245641},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.329461,\"nuts_mean\":0.00028956,\"nuts_sd\":0.196845,\"site\":\"s206\",\"svi_mean\":0.00648599,\"svi_sd\":0.199698},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0267296,\"nuts_mean\":-0.270502,\"nuts_sd\":0.212657,\"site\":\"s207\",\"svi_mean\":-0.266888,\"svi_sd\":0.205275},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.429909,\"nuts_mean\":-0.212824,\"nuts_sd\":0.217183,\"site\":\"s208\",\"svi_mean\":-0.217788,\"svi_sd\":0.217361},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.43206,\"nuts_mean\":0.249092,\"nuts_sd\":0.173524,\"site\":\"s209\",\"svi_mean\":0.250026,\"svi_sd\":0.170262},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.25765,\"nuts_mean\":0.693569,\"nuts_sd\":0.145686,\"site\":\"s210\",\"svi_mean\":0.691626,\"svi_sd\":0.144275},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.798521,\"nuts_mean\":0.257462,\"nuts_sd\":0.171783,\"site\":\"s211\",\"svi_mean\":0.250813,\"svi_sd\":0.170634},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.888363,\"nuts_mean\":0.6993,\"nuts_sd\":0.15205,\"site\":\"s212\",\"svi_mean\":0.699409,\"svi_sd\":0.151841},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.242376,\"nuts_mean\":0.28085,\"nuts_sd\":0.170132,\"site\":\"s213\",\"svi_mean\":0.281204,\"svi_sd\":0.168004},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.234664,\"nuts_mean\":0.166472,\"nuts_sd\":0.201748,\"site\":\"s214\",\"svi_mean\":0.168206,\"svi_sd\":0.199484},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.343692,\"nuts_mean\":0.292142,\"nuts_sd\":0.172205,\"site\":\"s215\",\"svi_mean\":0.288307,\"svi_sd\":0.166723},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.354874,\"nuts_mean\":0.252944,\"nuts_sd\":0.181351,\"site\":\"s216\",\"svi_mean\":0.256685,\"svi_sd\":0.171923},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.585815,\"nuts_mean\":-0.122342,\"nuts_sd\":0.209136,\"site\":\"s217\",\"svi_mean\":-0.123951,\"svi_sd\":0.208621},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.260121,\"nuts_mean\":0.422125,\"nuts_sd\":0.182087,\"site\":\"s218\",\"svi_mean\":0.423578,\"svi_sd\":0.181845},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.439625,\"nuts_mean\":-0.0571707,\"nuts_sd\":0.197473,\"site\":\"s219\",\"svi_mean\":-0.0530136,\"svi_sd\":0.191242},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0145345,\"nuts_mean\":1.07487,\"nuts_sd\":0.125641,\"site\":\"s220\",\"svi_mean\":1.07949,\"svi_sd\":0.120882},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.269509,\"nuts_mean\":-0.287833,\"nuts_sd\":0.210222,\"site\":\"s221\",\"svi_mean\":-0.287631,\"svi_sd\":0.207732},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.504619,\"nuts_mean\":0.841824,\"nuts_sd\":0.140476,\"site\":\"s222\",\"svi_mean\":0.844277,\"svi_sd\":0.134672},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0517826,\"nuts_mean\":0.28896,\"nuts_sd\":0.171754,\"site\":\"s223\",\"svi_mean\":0.29209,\"svi_sd\":0.16505},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.223453,\"nuts_mean\":-0.325088,\"nuts_sd\":0.216182,\"site\":\"s224\",\"svi_mean\":-0.330624,\"svi_sd\":0.216863},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.276666,\"nuts_mean\":0.196368,\"nuts_sd\":0.181359,\"site\":\"s225\",\"svi_mean\":0.196622,\"svi_sd\":0.175755},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.77696,\"nuts_mean\":-0.215557,\"nuts_sd\":0.212296,\"site\":\"s226\",\"svi_mean\":-0.21967,\"svi_sd\":0.212198},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.078179,\"nuts_mean\":-0.21301,\"nuts_sd\":0.210347,\"site\":\"s227\",\"svi_mean\":-0.207175,\"svi_sd\":0.209584},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0989201,\"nuts_mean\":0.215179,\"nuts_sd\":0.172593,\"site\":\"s228\",\"svi_mean\":0.231855,\"svi_sd\":0.169551},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.265329,\"nuts_mean\":0.195551,\"nuts_sd\":0.191341,\"site\":\"s229\",\"svi_mean\":0.192763,\"svi_sd\":0.19291},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.249153,\"nuts_mean\":-0.471655,\"nuts_sd\":0.256179,\"site\":\"s230\",\"svi_mean\":-0.4738,\"svi_sd\":0.260705},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.465495,\"nuts_mean\":0.0670487,\"nuts_sd\":0.184585,\"site\":\"s231\",\"svi_mean\":0.0658401,\"svi_sd\":0.180608},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.520493,\"nuts_mean\":0.274051,\"nuts_sd\":0.179587,\"site\":\"s232\",\"svi_mean\":0.274221,\"svi_sd\":0.173956},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.114966,\"nuts_mean\":-0.659378,\"nuts_sd\":0.249975,\"site\":\"s233\",\"svi_mean\":-0.655261,\"svi_sd\":0.244226},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.402965,\"nuts_mean\":-0.180899,\"nuts_sd\":0.210626,\"site\":\"s234\",\"svi_mean\":-0.190177,\"svi_sd\":0.207357},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.000487395,\"nuts_mean\":-0.438454,\"nuts_sd\":0.233936,\"site\":\"s235\",\"svi_mean\":-0.442132,\"svi_sd\":0.226319},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0755706,\"nuts_mean\":0.39161,\"nuts_sd\":0.164784,\"site\":\"s236\",\"svi_mean\":0.388695,\"svi_sd\":0.164862},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0771007,\"nuts_mean\":-0.565717,\"nuts_sd\":0.243341,\"site\":\"s237\",\"svi_mean\":-0.573664,\"svi_sd\":0.239625},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.776319,\"nuts_mean\":0.187355,\"nuts_sd\":0.177908,\"site\":\"s238\",\"svi_mean\":0.190325,\"svi_sd\":0.181252},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.57767,\"nuts_mean\":-0.152249,\"nuts_sd\":0.204829,\"site\":\"s239\",\"svi_mean\":-0.152175,\"svi_sd\":0.200548},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.458583,\"nuts_mean\":-0.699194,\"nuts_sd\":0.249266,\"site\":\"s240\",\"svi_mean\":-0.686834,\"svi_sd\":0.252321},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.810908,\"nuts_mean\":-0.244722,\"nuts_sd\":0.212981,\"site\":\"s241\",\"svi_mean\":-0.246745,\"svi_sd\":0.207828},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.309042,\"nuts_mean\":0.56336,\"nuts_sd\":0.154719,\"site\":\"s242\",\"svi_mean\":0.558852,\"svi_sd\":0.153369},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0110713,\"nuts_mean\":1.04818,\"nuts_sd\":0.138342,\"site\":\"s243\",\"svi_mean\":1.04622,\"svi_sd\":0.131403},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.473141,\"nuts_mean\":-0.220375,\"nuts_sd\":0.208404,\"site\":\"s244\",\"svi_mean\":-0.214343,\"svi_sd\":0.203698},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.534868,\"nuts_mean\":0.36942,\"nuts_sd\":0.163919,\"site\":\"s245\",\"svi_mean\":0.371923,\"svi_sd\":0.166095},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.457246,\"nuts_mean\":0.314941,\"nuts_sd\":0.172288,\"site\":\"s246\",\"svi_mean\":0.314242,\"svi_sd\":0.167557},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.685507,\"nuts_mean\":-0.0653699,\"nuts_sd\":0.197231,\"site\":\"s247\",\"svi_mean\":-0.0632022,\"svi_sd\":0.201838},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.446737,\"nuts_mean\":0.321634,\"nuts_sd\":0.165253,\"site\":\"s248\",\"svi_mean\":0.319142,\"svi_sd\":0.161657},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.00797366,\"nuts_mean\":0.234323,\"nuts_sd\":0.175403,\"site\":\"s249\",\"svi_mean\":0.24054,\"svi_sd\":0.170809},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.263205,\"nuts_mean\":0.082604,\"nuts_sd\":0.203014,\"site\":\"s250\",\"svi_mean\":0.087861,\"svi_sd\":0.198685},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.458373,\"nuts_mean\":0.367152,\"nuts_sd\":0.176693,\"site\":\"s251\",\"svi_mean\":0.368076,\"svi_sd\":0.179933},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.298584,\"nuts_mean\":0.502185,\"nuts_sd\":0.157302,\"site\":\"s252\",\"svi_mean\":0.499522,\"svi_sd\":0.159664},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.212104,\"nuts_mean\":-0.458545,\"nuts_sd\":0.236626,\"site\":\"s253\",\"svi_mean\":-0.457215,\"svi_sd\":0.237311},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.766322,\"nuts_mean\":-0.196641,\"nuts_sd\":0.212355,\"site\":\"s254\",\"svi_mean\":-0.198727,\"svi_sd\":0.213577},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.411433,\"nuts_mean\":-0.56139,\"nuts_sd\":0.238371,\"site\":\"s255\",\"svi_mean\":-0.566433,\"svi_sd\":0.234541},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.135124,\"nuts_mean\":-0.63408,\"nuts_sd\":0.24425,\"site\":\"s256\",\"svi_mean\":-0.627864,\"svi_sd\":0.238536},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.139217,\"nuts_mean\":-0.0870971,\"nuts_sd\":0.192936,\"site\":\"s257\",\"svi_mean\":-0.0879296,\"svi_sd\":0.187761},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.484095,\"nuts_mean\":-0.298381,\"nuts_sd\":0.244943,\"site\":\"s258\",\"svi_mean\":-0.296046,\"svi_sd\":0.248577},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.910665,\"nuts_mean\":0.431933,\"nuts_sd\":0.197544,\"site\":\"s259\",\"svi_mean\":0.430939,\"svi_sd\":0.196981},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.286304,\"nuts_mean\":-0.501197,\"nuts_sd\":0.236331,\"site\":\"s260\",\"svi_mean\":-0.494809,\"svi_sd\":0.229146},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.516361,\"nuts_mean\":0.273962,\"nuts_sd\":0.169714,\"site\":\"s261\",\"svi_mean\":0.273564,\"svi_sd\":0.167609},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.447757,\"nuts_mean\":0.522809,\"nuts_sd\":0.17275,\"site\":\"s262\",\"svi_mean\":0.518236,\"svi_sd\":0.168895},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0161274,\"nuts_mean\":0.215097,\"nuts_sd\":0.178097,\"site\":\"s263\",\"svi_mean\":0.220017,\"svi_sd\":0.17503},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.468115,\"nuts_mean\":0.444344,\"nuts_sd\":0.161044,\"site\":\"s264\",\"svi_mean\":0.439683,\"svi_sd\":0.15821},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.214448,\"nuts_mean\":-1.11016,\"nuts_sd\":0.290482,\"site\":\"s265\",\"svi_mean\":-1.10297,\"svi_sd\":0.289045},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.32401,\"nuts_mean\":-0.216203,\"nuts_sd\":0.209224,\"site\":\"s266\",\"svi_mean\":-0.210379,\"svi_sd\":0.203735},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.519739,\"nuts_mean\":-0.00717992,\"nuts_sd\":0.191429,\"site\":\"s267\",\"svi_mean\":0.00120584,\"svi_sd\":0.194412},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.727341,\"nuts_mean\":-0.701734,\"nuts_sd\":0.247917,\"site\":\"s268\",\"svi_mean\":-0.686538,\"svi_sd\":0.242168},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.167093,\"nuts_mean\":0.0511902,\"nuts_sd\":0.18353,\"site\":\"s269\",\"svi_mean\":0.0521529,\"svi_sd\":0.185041},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0830427,\"nuts_mean\":-0.0881406,\"nuts_sd\":0.218819,\"site\":\"s270\",\"svi_mean\":-0.0952207,\"svi_sd\":0.217335},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.101088,\"nuts_mean\":-0.0104023,\"nuts_sd\":0.194792,\"site\":\"s271\",\"svi_mean\":-0.00594949,\"svi_sd\":0.190092},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.378044,\"nuts_mean\":0.5987,\"nuts_sd\":0.176133,\"site\":\"s272\",\"svi_mean\":0.600698,\"svi_sd\":0.168797},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.61898,\"nuts_mean\":1.04414,\"nuts_sd\":0.139112,\"site\":\"s273\",\"svi_mean\":1.04826,\"svi_sd\":0.13878},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.55189,\"nuts_mean\":0.918981,\"nuts_sd\":0.131468,\"site\":\"s274\",\"svi_mean\":0.916701,\"svi_sd\":0.129152},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.123463,\"nuts_mean\":0.298782,\"nuts_sd\":0.175009,\"site\":\"s275\",\"svi_mean\":0.300688,\"svi_sd\":0.167794},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.03453,\"nuts_mean\":-0.459254,\"nuts_sd\":0.228212,\"site\":\"s276\",\"svi_mean\":-0.461455,\"svi_sd\":0.233034},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.382763,\"nuts_mean\":-0.621455,\"nuts_sd\":0.241334,\"site\":\"s277\",\"svi_mean\":-0.609019,\"svi_sd\":0.237085},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.434474,\"nuts_mean\":0.516722,\"nuts_sd\":0.152861,\"site\":\"s278\",\"svi_mean\":0.515255,\"svi_sd\":0.15175},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.17506,\"nuts_mean\":0.395789,\"nuts_sd\":0.172651,\"site\":\"s279\",\"svi_mean\":0.396064,\"svi_sd\":0.171403},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.197822,\"nuts_mean\":0.424773,\"nuts_sd\":0.164531,\"site\":\"s280\",\"svi_mean\":0.425561,\"svi_sd\":0.164024},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.147641,\"nuts_mean\":-0.77763,\"nuts_sd\":0.259117,\"site\":\"s281\",\"svi_mean\":-0.775699,\"svi_sd\":0.257198},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.82489,\"nuts_mean\":-0.864394,\"nuts_sd\":0.282354,\"site\":\"s282\",\"svi_mean\":-0.860231,\"svi_sd\":0.281765},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.43431,\"nuts_mean\":0.843187,\"nuts_sd\":0.146473,\"site\":\"s283\",\"svi_mean\":0.842419,\"svi_sd\":0.145979},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.178136,\"nuts_mean\":-0.51477,\"nuts_sd\":0.233417,\"site\":\"s284\",\"svi_mean\":-0.509034,\"svi_sd\":0.22976},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.345994,\"nuts_mean\":0.00353801,\"nuts_sd\":0.190757,\"site\":\"s285\",\"svi_mean\":0.00982531,\"svi_sd\":0.193169},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.308441,\"nuts_mean\":0.170966,\"nuts_sd\":0.180277,\"site\":\"s286\",\"svi_mean\":0.169084,\"svi_sd\":0.17639},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.203612,\"nuts_mean\":0.85079,\"nuts_sd\":0.135008,\"site\":\"s287\",\"svi_mean\":0.84743,\"svi_sd\":0.133192},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.677893,\"nuts_mean\":-0.0864763,\"nuts_sd\":0.214362,\"site\":\"s288\",\"svi_mean\":-0.0910691,\"svi_sd\":0.213949},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.288823,\"nuts_mean\":0.196393,\"nuts_sd\":0.187182,\"site\":\"s289\",\"svi_mean\":0.20015,\"svi_sd\":0.184006},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.244794,\"nuts_mean\":0.208237,\"nuts_sd\":0.177391,\"site\":\"s290\",\"svi_mean\":0.207977,\"svi_sd\":0.182409},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.113074,\"nuts_mean\":0.261,\"nuts_sd\":0.192427,\"site\":\"s291\",\"svi_mean\":0.259004,\"svi_sd\":0.193514},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.33932,\"nuts_mean\":0.227945,\"nuts_sd\":0.177467,\"site\":\"s292\",\"svi_mean\":0.225678,\"svi_sd\":0.177411},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.158519,\"nuts_mean\":0.441262,\"nuts_sd\":0.165213,\"site\":\"s293\",\"svi_mean\":0.440566,\"svi_sd\":0.164741},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.038441,\"nuts_mean\":0.0160187,\"nuts_sd\":0.194818,\"site\":\"s294\",\"svi_mean\":0.0177764,\"svi_sd\":0.195498},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.225604,\"nuts_mean\":0.153989,\"nuts_sd\":0.192233,\"site\":\"s295\",\"svi_mean\":0.139811,\"svi_sd\":0.189392},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.163964,\"nuts_mean\":0.353276,\"nuts_sd\":0.182495,\"site\":\"s296\",\"svi_mean\":0.361472,\"svi_sd\":0.176147},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.02171,\"nuts_mean\":-0.286687,\"nuts_sd\":0.227228,\"site\":\"s297\",\"svi_mean\":-0.283046,\"svi_sd\":0.222237},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0658127,\"nuts_mean\":0.565024,\"nuts_sd\":0.156084,\"site\":\"s298\",\"svi_mean\":0.562744,\"svi_sd\":0.156293},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.795749,\"nuts_mean\":-0.195861,\"nuts_sd\":0.213238,\"site\":\"s299\",\"svi_mean\":-0.195704,\"svi_sd\":0.211815},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.216624,\"nuts_mean\":-0.597884,\"nuts_sd\":0.248913,\"site\":\"s300\",\"svi_mean\":-0.596899,\"svi_sd\":0.240339},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.010686,\"nuts_mean\":0.246428,\"nuts_sd\":0.172159,\"site\":\"s301\",\"svi_mean\":0.251267,\"svi_sd\":0.172114},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.231289,\"nuts_mean\":-0.258027,\"nuts_sd\":0.222335,\"site\":\"s302\",\"svi_mean\":-0.248101,\"svi_sd\":0.211565},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.544499,\"nuts_mean\":-0.359611,\"nuts_sd\":0.224633,\"site\":\"s303\",\"svi_mean\":-0.356901,\"svi_sd\":0.224006},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.120542,\"nuts_mean\":-0.0302623,\"nuts_sd\":0.200959,\"site\":\"s304\",\"svi_mean\":-0.026863,\"svi_sd\":0.202228},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.144204,\"nuts_mean\":-0.129845,\"nuts_sd\":0.217044,\"site\":\"s305\",\"svi_mean\":-0.13347,\"svi_sd\":0.206468},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0411291,\"nuts_mean\":0.176637,\"nuts_sd\":0.210689,\"site\":\"s306\",\"svi_mean\":0.167965,\"svi_sd\":0.208562},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.495002,\"nuts_mean\":-0.0216463,\"nuts_sd\":0.224901,\"site\":\"s307\",\"svi_mean\":-0.0206133,\"svi_sd\":0.221304},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.240215,\"nuts_mean\":0.116596,\"nuts_sd\":0.183252,\"site\":\"s308\",\"svi_mean\":0.119509,\"svi_sd\":0.176055},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.223233,\"nuts_mean\":0.106814,\"nuts_sd\":0.179441,\"site\":\"s309\",\"svi_mean\":0.11207,\"svi_sd\":0.175261},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.238759,\"nuts_mean\":-0.805538,\"nuts_sd\":0.268694,\"site\":\"s310\",\"svi_mean\":-0.798005,\"svi_sd\":0.265766},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.457367,\"nuts_mean\":0.06461,\"nuts_sd\":0.236978,\"site\":\"s311\",\"svi_mean\":0.0581972,\"svi_sd\":0.222142},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.192523,\"nuts_mean\":0.853465,\"nuts_sd\":0.135297,\"site\":\"s312\",\"svi_mean\":0.855436,\"svi_sd\":0.132824},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.167042,\"nuts_mean\":-0.0572927,\"nuts_sd\":0.199265,\"site\":\"s313\",\"svi_mean\":-0.0558496,\"svi_sd\":0.190467},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.832631,\"nuts_mean\":0.540633,\"nuts_sd\":0.158389,\"site\":\"s314\",\"svi_mean\":0.545232,\"svi_sd\":0.160333},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.366908,\"nuts_mean\":0.155933,\"nuts_sd\":0.184564,\"site\":\"s315\",\"svi_mean\":0.152455,\"svi_sd\":0.176852},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.490386,\"nuts_mean\":-0.216388,\"nuts_sd\":0.208129,\"site\":\"s316\",\"svi_mean\":-0.209972,\"svi_sd\":0.204097},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.895109,\"nuts_mean\":-0.215032,\"nuts_sd\":0.215442,\"site\":\"s317\",\"svi_mean\":-0.214795,\"svi_sd\":0.20761},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.377165,\"nuts_mean\":-0.0670636,\"nuts_sd\":0.195697,\"site\":\"s318\",\"svi_mean\":-0.0617634,\"svi_sd\":0.195542},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.462063,\"nuts_mean\":-0.00705521,\"nuts_sd\":0.215623,\"site\":\"s319\",\"svi_mean\":-0.0131073,\"svi_sd\":0.21895},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.100698,\"nuts_mean\":0.434956,\"nuts_sd\":0.166926,\"site\":\"s320\",\"svi_mean\":0.432826,\"svi_sd\":0.167033},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.43908,\"nuts_mean\":0.663771,\"nuts_sd\":0.149959,\"site\":\"s321\",\"svi_mean\":0.662067,\"svi_sd\":0.148604},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0719551,\"nuts_mean\":0.220599,\"nuts_sd\":0.185168,\"site\":\"s322\",\"svi_mean\":0.221124,\"svi_sd\":0.183051},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.200751,\"nuts_mean\":0.487171,\"nuts_sd\":0.158274,\"site\":\"s323\",\"svi_mean\":0.492935,\"svi_sd\":0.153402},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.326276,\"nuts_mean\":-0.177292,\"nuts_sd\":0.28975,\"site\":\"s324\",\"svi_mean\":-0.167913,\"svi_sd\":0.285834},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.501165,\"nuts_mean\":0.0546153,\"nuts_sd\":0.18377,\"site\":\"s325\",\"svi_mean\":0.0583392,\"svi_sd\":0.181006},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.486645,\"nuts_mean\":-0.779758,\"nuts_sd\":0.254191,\"site\":\"s326\",\"svi_mean\":-0.770846,\"svi_sd\":0.26379},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.402616,\"nuts_mean\":-0.489729,\"nuts_sd\":0.231018,\"site\":\"s327\",\"svi_mean\":-0.494716,\"svi_sd\":0.224035},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.478997,\"nuts_mean\":-0.341671,\"nuts_sd\":0.221347,\"site\":\"s328\",\"svi_mean\":-0.333395,\"svi_sd\":0.218329},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.784055,\"nuts_mean\":-0.0965703,\"nuts_sd\":0.206473,\"site\":\"s329\",\"svi_mean\":-0.0957505,\"svi_sd\":0.20417},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0962954,\"nuts_mean\":-0.29202,\"nuts_sd\":0.215676,\"site\":\"s330\",\"svi_mean\":-0.281556,\"svi_sd\":0.215119},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.573984,\"nuts_mean\":-0.466311,\"nuts_sd\":0.229074,\"site\":\"s331\",\"svi_mean\":-0.457198,\"svi_sd\":0.229934},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.884392,\"nuts_mean\":-0.238669,\"nuts_sd\":0.215812,\"site\":\"s332\",\"svi_mean\":-0.237462,\"svi_sd\":0.207505},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.207024,\"nuts_mean\":-0.65996,\"nuts_sd\":0.251502,\"site\":\"s333\",\"svi_mean\":-0.655283,\"svi_sd\":0.252544},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.512946,\"nuts_mean\":0.842828,\"nuts_sd\":0.144353,\"site\":\"s334\",\"svi_mean\":0.845236,\"svi_sd\":0.146798},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.364536,\"nuts_mean\":-0.00262631,\"nuts_sd\":0.191584,\"site\":\"s335\",\"svi_mean\":0.00425178,\"svi_sd\":0.189312},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.56283,\"nuts_mean\":-0.395358,\"nuts_sd\":0.240227,\"site\":\"s336\",\"svi_mean\":-0.390109,\"svi_sd\":0.240964},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.832918,\"nuts_mean\":-0.705558,\"nuts_sd\":0.24874,\"site\":\"s337\",\"svi_mean\":-0.704455,\"svi_sd\":0.249495},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.119549,\"nuts_mean\":-0.985827,\"nuts_sd\":0.275912,\"site\":\"s338\",\"svi_mean\":-0.98886,\"svi_sd\":0.279908},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0709199,\"nuts_mean\":-0.716731,\"nuts_sd\":0.272585,\"site\":\"s339\",\"svi_mean\":-0.702561,\"svi_sd\":0.267651},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.00583134,\"nuts_mean\":-0.243433,\"nuts_sd\":0.216398,\"site\":\"s340\",\"svi_mean\":-0.238521,\"svi_sd\":0.212479},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0733642,\"nuts_mean\":0.935933,\"nuts_sd\":0.130701,\"site\":\"s341\",\"svi_mean\":0.937201,\"svi_sd\":0.128899},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0489546,\"nuts_mean\":0.938403,\"nuts_sd\":0.137106,\"site\":\"s342\",\"svi_mean\":0.9404,\"svi_sd\":0.135913},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.753078,\"nuts_mean\":0.331647,\"nuts_sd\":0.172221,\"site\":\"s343\",\"svi_mean\":0.328177,\"svi_sd\":0.17212},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.987166,\"nuts_mean\":-0.286772,\"nuts_sd\":0.21828,\"site\":\"s344\",\"svi_mean\":-0.283338,\"svi_sd\":0.212805},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.02924,\"nuts_mean\":0.00262214,\"nuts_sd\":0.19276,\"site\":\"s345\",\"svi_mean\":0.00917641,\"svi_sd\":0.190245},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.454037,\"nuts_mean\":-0.563752,\"nuts_sd\":0.235152,\"site\":\"s346\",\"svi_mean\":-0.570429,\"svi_sd\":0.232061},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.48482,\"nuts_mean\":0.325009,\"nuts_sd\":0.175341,\"site\":\"s347\",\"svi_mean\":0.327468,\"svi_sd\":0.171246},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.608683,\"nuts_mean\":0.70921,\"nuts_sd\":0.142291,\"site\":\"s348\",\"svi_mean\":0.710412,\"svi_sd\":0.13708},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.715523,\"nuts_mean\":-0.110817,\"nuts_sd\":0.205009,\"site\":\"s349\",\"svi_mean\":-0.106501,\"svi_sd\":0.205544},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.565395,\"nuts_mean\":-0.254333,\"nuts_sd\":0.215899,\"site\":\"s350\",\"svi_mean\":-0.241869,\"svi_sd\":0.207122},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.761281,\"nuts_mean\":0.701599,\"nuts_sd\":0.144995,\"site\":\"s351\",\"svi_mean\":0.704636,\"svi_sd\":0.147772},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.990725,\"nuts_mean\":-0.0604251,\"nuts_sd\":0.19923,\"site\":\"s352\",\"svi_mean\":-0.0598302,\"svi_sd\":0.203357},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":1.09204,\"nuts_mean\":-0.488793,\"nuts_sd\":0.228147,\"site\":\"s353\",\"svi_mean\":-0.483746,\"svi_sd\":0.222493},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-1.28535,\"nuts_mean\":-0.11436,\"nuts_sd\":0.210402,\"site\":\"s354\",\"svi_mean\":-0.11195,\"svi_sd\":0.209513},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.164153,\"nuts_mean\":-0.80545,\"nuts_sd\":0.266931,\"site\":\"s355\",\"svi_mean\":-0.804469,\"svi_sd\":0.273918},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.330955,\"nuts_mean\":-0.268467,\"nuts_sd\":0.217482,\"site\":\"s356\",\"svi_mean\":-0.27073,\"svi_sd\":0.216045},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.641855,\"nuts_mean\":-0.491423,\"nuts_sd\":0.229709,\"site\":\"s357\",\"svi_mean\":-0.491366,\"svi_sd\":0.225754},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.62901,\"nuts_mean\":-0.307919,\"nuts_sd\":0.220866,\"site\":\"s358\",\"svi_mean\":-0.3008,\"svi_sd\":0.219758},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.105548,\"nuts_mean\":0.134919,\"nuts_sd\":0.187819,\"site\":\"s359\",\"svi_mean\":0.132705,\"svi_sd\":0.184857},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.508554,\"nuts_mean\":0.602236,\"nuts_sd\":0.15294,\"site\":\"s360\",\"svi_mean\":0.605843,\"svi_sd\":0.152978},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.176208,\"nuts_mean\":-0.58161,\"nuts_sd\":0.242533,\"site\":\"s361\",\"svi_mean\":-0.571948,\"svi_sd\":0.235778},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.573698,\"nuts_mean\":-0.234945,\"nuts_sd\":0.210071,\"site\":\"s362\",\"svi_mean\":-0.236078,\"svi_sd\":0.208668},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0590756,\"nuts_mean\":-0.0829873,\"nuts_sd\":0.199825,\"site\":\"s363\",\"svi_mean\":-0.0796624,\"svi_sd\":0.193586},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.403041,\"nuts_mean\":-0.13316,\"nuts_sd\":0.212851,\"site\":\"s364\",\"svi_mean\":-0.131775,\"svi_sd\":0.206823},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.311355,\"nuts_mean\":-0.141614,\"nuts_sd\":0.237981,\"site\":\"s365\",\"svi_mean\":-0.137725,\"svi_sd\":0.238306},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.304927,\"nuts_mean\":-0.341052,\"nuts_sd\":0.217365,\"site\":\"s366\",\"svi_mean\":-0.332673,\"svi_sd\":0.212384},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.660127,\"nuts_mean\":0.115916,\"nuts_sd\":0.182143,\"site\":\"s367\",\"svi_mean\":0.107311,\"svi_sd\":0.181005},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.310494,\"nuts_mean\":0.240826,\"nuts_sd\":0.182407,\"site\":\"s368\",\"svi_mean\":0.238754,\"svi_sd\":0.183194},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.557689,\"nuts_mean\":0.75488,\"nuts_sd\":0.142544,\"site\":\"s369\",\"svi_mean\":0.758161,\"svi_sd\":0.143283},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0502256,\"nuts_mean\":1.01206,\"nuts_sd\":0.140316,\"site\":\"s370\",\"svi_mean\":1.01624,\"svi_sd\":0.135048},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.299444,\"nuts_mean\":-0.444386,\"nuts_sd\":0.230572,\"site\":\"s371\",\"svi_mean\":-0.44711,\"svi_sd\":0.225285},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.237266,\"nuts_mean\":-0.129854,\"nuts_sd\":0.368299,\"site\":\"s372\",\"svi_mean\":-0.139953,\"svi_sd\":0.352613},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.188227,\"nuts_mean\":0.237036,\"nuts_sd\":0.17729,\"site\":\"s373\",\"svi_mean\":0.240278,\"svi_sd\":0.172237},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.536966,\"nuts_mean\":0.148245,\"nuts_sd\":0.187619,\"site\":\"s374\",\"svi_mean\":0.154236,\"svi_sd\":0.187102},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0208528,\"nuts_mean\":-0.141652,\"nuts_sd\":0.204997,\"site\":\"s375\",\"svi_mean\":-0.140936,\"svi_sd\":0.199318},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.336791,\"nuts_mean\":-0.260817,\"nuts_sd\":0.215421,\"site\":\"s376\",\"svi_mean\":-0.251948,\"svi_sd\":0.2158},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0081649,\"nuts_mean\":1.04007,\"nuts_sd\":0.128789,\"site\":\"s377\",\"svi_mean\":1.03712,\"svi_sd\":0.12348},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-1.36946,\"nuts_mean\":-0.341364,\"nuts_sd\":0.242086,\"site\":\"s378\",\"svi_mean\":-0.347001,\"svi_sd\":0.235481},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0447745,\"nuts_mean\":0.625148,\"nuts_sd\":0.15375,\"site\":\"s379\",\"svi_mean\":0.630128,\"svi_sd\":0.157202},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.263555,\"nuts_mean\":0.203105,\"nuts_sd\":0.186957,\"site\":\"s380\",\"svi_mean\":0.209065,\"svi_sd\":0.180256},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.383495,\"nuts_mean\":-0.0452981,\"nuts_sd\":0.206546,\"site\":\"s381\",\"svi_mean\":-0.053957,\"svi_sd\":0.205372},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.00199047,\"nuts_mean\":-0.131455,\"nuts_sd\":0.208829,\"site\":\"s382\",\"svi_mean\":-0.126341,\"svi_sd\":0.196488},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.390327,\"nuts_mean\":-0.215367,\"nuts_sd\":0.278421,\"site\":\"s383\",\"svi_mean\":-0.203552,\"svi_sd\":0.269576},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0728147,\"nuts_mean\":0.55786,\"nuts_sd\":0.157035,\"site\":\"s384\",\"svi_mean\":0.549118,\"svi_sd\":0.149279},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.0311538,\"nuts_mean\":0.605297,\"nuts_sd\":0.151391,\"site\":\"s385\",\"svi_mean\":0.600411,\"svi_sd\":0.149455},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.274663,\"nuts_mean\":0.0892213,\"nuts_sd\":0.191675,\"site\":\"s386\",\"svi_mean\":0.0883495,\"svi_sd\":0.18908},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.401422,\"nuts_mean\":-0.382516,\"nuts_sd\":0.224404,\"site\":\"s387\",\"svi_mean\":-0.371082,\"svi_sd\":0.225077},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.725988,\"nuts_mean\":0.691415,\"nuts_sd\":0.144477,\"site\":\"s388\",\"svi_mean\":0.686955,\"svi_sd\":0.141634},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.323904,\"nuts_mean\":0.376711,\"nuts_sd\":0.167915,\"site\":\"s389\",\"svi_mean\":0.369876,\"svi_sd\":0.168184},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.257998,\"nuts_mean\":0.132038,\"nuts_sd\":0.186011,\"site\":\"s390\",\"svi_mean\":0.130453,\"svi_sd\":0.179414},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.138576,\"nuts_mean\":0.439954,\"nuts_sd\":0.207629,\"site\":\"s391\",\"svi_mean\":0.445453,\"svi_sd\":0.201618},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.413481,\"nuts_mean\":-0.0730866,\"nuts_sd\":0.201775,\"site\":\"s392\",\"svi_mean\":-0.0668764,\"svi_sd\":0.201641},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.414171,\"nuts_mean\":-0.357688,\"nuts_sd\":0.225004,\"site\":\"s393\",\"svi_mean\":-0.35183,\"svi_sd\":0.224827},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.370866,\"nuts_mean\":0.6876,\"nuts_sd\":0.147817,\"site\":\"s394\",\"svi_mean\":0.686812,\"svi_sd\":0.145182},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.346523,\"nuts_mean\":-0.525289,\"nuts_sd\":0.231954,\"site\":\"s395\",\"svi_mean\":-0.523649,\"svi_sd\":0.235094},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.549355,\"nuts_mean\":-0.379269,\"nuts_sd\":0.223392,\"site\":\"s396\",\"svi_mean\":-0.378933,\"svi_sd\":0.218105},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-1.04111,\"nuts_mean\":-0.0270999,\"nuts_sd\":0.191827,\"site\":\"s397\",\"svi_mean\":-0.0284551,\"svi_sd\":0.193753},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":0.0969479,\"nuts_mean\":0.645858,\"nuts_sd\":0.14812,\"site\":\"s398\",\"svi_mean\":0.648948,\"svi_sd\":0.148569},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.84108,\"nuts_mean\":0.957045,\"nuts_sd\":0.141724,\"site\":\"s399\",\"svi_mean\":0.955046,\"svi_sd\":0.134411},{\"__src\":\"_s1\",\"effect\":\"site intercept deviation\",\"generating\":-0.449037,\"nuts_mean\":-0.698051,\"nuts_sd\":0.25516,\"site\":\"s400\",\"svi_mean\":-0.688539,\"svi_sd\":0.250412},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.121455,\"nuts_mean\":-0.14699,\"nuts_sd\":0.196342,\"site\":\"s1\",\"svi_mean\":-0.15152,\"svi_sd\":0.193327},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.525441,\"nuts_mean\":-0.101171,\"nuts_sd\":0.19727,\"site\":\"s2\",\"svi_mean\":-0.103819,\"svi_sd\":0.189993},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.151176,\"nuts_mean\":-0.0845037,\"nuts_sd\":0.19026,\"site\":\"s3\",\"svi_mean\":-0.0864305,\"svi_sd\":0.190453},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0657484,\"nuts_mean\":0.0244336,\"nuts_sd\":0.201734,\"site\":\"s4\",\"svi_mean\":0.019388,\"svi_sd\":0.200082},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.1567,\"nuts_mean\":-0.255923,\"nuts_sd\":0.149815,\"site\":\"s5\",\"svi_mean\":-0.251457,\"svi_sd\":0.14929},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.245778,\"nuts_mean\":0.474107,\"nuts_sd\":0.106926,\"site\":\"s6\",\"svi_mean\":0.475138,\"svi_sd\":0.104036},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0354418,\"nuts_mean\":-0.0951594,\"nuts_sd\":0.129832,\"site\":\"s7\",\"svi_mean\":-0.0976507,\"svi_sd\":0.130319},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.23448,\"nuts_mean\":-0.194362,\"nuts_sd\":0.217778,\"site\":\"s8\",\"svi_mean\":-0.188804,\"svi_sd\":0.215496},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.023117,\"nuts_mean\":-0.362744,\"nuts_sd\":0.203092,\"site\":\"s9\",\"svi_mean\":-0.351248,\"svi_sd\":0.198055},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00449957,\"nuts_mean\":-0.033854,\"nuts_sd\":0.135011,\"site\":\"s10\",\"svi_mean\":-0.0398391,\"svi_sd\":0.134434},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.237132,\"nuts_mean\":0.0294885,\"nuts_sd\":0.136466,\"site\":\"s11\",\"svi_mean\":0.0262736,\"svi_sd\":0.134454},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.138192,\"nuts_mean\":-0.0277833,\"nuts_sd\":0.214729,\"site\":\"s12\",\"svi_mean\":-0.0302302,\"svi_sd\":0.217046},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.332247,\"nuts_mean\":-0.206513,\"nuts_sd\":0.167154,\"site\":\"s13\",\"svi_mean\":-0.205726,\"svi_sd\":0.170056},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.568121,\"nuts_mean\":-0.15994,\"nuts_sd\":0.2039,\"site\":\"s14\",\"svi_mean\":-0.164542,\"svi_sd\":0.199706},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.449491,\"nuts_mean\":-0.0374957,\"nuts_sd\":0.179958,\"site\":\"s15\",\"svi_mean\":-0.0391193,\"svi_sd\":0.176527},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0487755,\"nuts_mean\":-0.00788107,\"nuts_sd\":0.237456,\"site\":\"s16\",\"svi_mean\":0.000105549,\"svi_sd\":0.240393},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.441697,\"nuts_mean\":-0.0621051,\"nuts_sd\":0.110223,\"site\":\"s17\",\"svi_mean\":-0.0566498,\"svi_sd\":0.11176},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0677361,\"nuts_mean\":-0.0899992,\"nuts_sd\":0.243799,\"site\":\"s18\",\"svi_mean\":-0.0904744,\"svi_sd\":0.247796},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.110009,\"nuts_mean\":-0.207503,\"nuts_sd\":0.132499,\"site\":\"s19\",\"svi_mean\":-0.197944,\"svi_sd\":0.131371},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.383565,\"nuts_mean\":-0.153598,\"nuts_sd\":0.165526,\"site\":\"s20\",\"svi_mean\":-0.156821,\"svi_sd\":0.171752},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.230791,\"nuts_mean\":-0.27548,\"nuts_sd\":0.228012,\"site\":\"s21\",\"svi_mean\":-0.278785,\"svi_sd\":0.227153},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.199072,\"nuts_mean\":0.0400801,\"nuts_sd\":0.18766,\"site\":\"s22\",\"svi_mean\":0.0368199,\"svi_sd\":0.193421},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0288401,\"nuts_mean\":0.0122386,\"nuts_sd\":0.18962,\"site\":\"s23\",\"svi_mean\":0.017461,\"svi_sd\":0.189803},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.219973,\"nuts_mean\":-0.0662646,\"nuts_sd\":0.133422,\"site\":\"s24\",\"svi_mean\":-0.0650409,\"svi_sd\":0.135685},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.36219,\"nuts_mean\":-0.073966,\"nuts_sd\":0.0985614,\"site\":\"s25\",\"svi_mean\":-0.073102,\"svi_sd\":0.0960665},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.345111,\"nuts_mean\":-0.0519322,\"nuts_sd\":0.191655,\"site\":\"s26\",\"svi_mean\":-0.0525013,\"svi_sd\":0.191313},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.346785,\"nuts_mean\":-0.247994,\"nuts_sd\":0.17239,\"site\":\"s27\",\"svi_mean\":-0.24847,\"svi_sd\":0.172976},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.700726,\"nuts_mean\":-0.213094,\"nuts_sd\":0.158502,\"site\":\"s28\",\"svi_mean\":-0.209337,\"svi_sd\":0.156904},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.428819,\"nuts_mean\":-0.082035,\"nuts_sd\":0.191021,\"site\":\"s29\",\"svi_mean\":-0.0846025,\"svi_sd\":0.190535},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0880806,\"nuts_mean\":-0.21915,\"nuts_sd\":0.206491,\"site\":\"s30\",\"svi_mean\":-0.212884,\"svi_sd\":0.204406},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.57229,\"nuts_mean\":0.202283,\"nuts_sd\":0.176438,\"site\":\"s31\",\"svi_mean\":0.197357,\"svi_sd\":0.175545},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.259907,\"nuts_mean\":-0.0449846,\"nuts_sd\":0.203277,\"site\":\"s32\",\"svi_mean\":-0.0473871,\"svi_sd\":0.204357},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.15345,\"nuts_mean\":-0.174486,\"nuts_sd\":0.215486,\"site\":\"s33\",\"svi_mean\":-0.171545,\"svi_sd\":0.215056},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.248007,\"nuts_mean\":-0.0854381,\"nuts_sd\":0.148625,\"site\":\"s34\",\"svi_mean\":-0.0837144,\"svi_sd\":0.14289},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0978966,\"nuts_mean\":0.23318,\"nuts_sd\":0.207736,\"site\":\"s35\",\"svi_mean\":0.228173,\"svi_sd\":0.202694},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.134371,\"nuts_mean\":0.275298,\"nuts_sd\":0.113522,\"site\":\"s36\",\"svi_mean\":0.279702,\"svi_sd\":0.112189},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.117394,\"nuts_mean\":-0.352962,\"nuts_sd\":0.138208,\"site\":\"s37\",\"svi_mean\":-0.353586,\"svi_sd\":0.140762},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0929571,\"nuts_mean\":0.213178,\"nuts_sd\":0.158726,\"site\":\"s38\",\"svi_mean\":0.210613,\"svi_sd\":0.155512},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.154345,\"nuts_mean\":0.111298,\"nuts_sd\":0.242582,\"site\":\"s39\",\"svi_mean\":0.106454,\"svi_sd\":0.240868},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0244275,\"nuts_mean\":-0.084136,\"nuts_sd\":0.20545,\"site\":\"s40\",\"svi_mean\":-0.0764942,\"svi_sd\":0.202125},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.228834,\"nuts_mean\":0.329177,\"nuts_sd\":0.180961,\"site\":\"s41\",\"svi_mean\":0.328344,\"svi_sd\":0.183799},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.156475,\"nuts_mean\":-0.179587,\"nuts_sd\":0.208735,\"site\":\"s42\",\"svi_mean\":-0.186649,\"svi_sd\":0.203617},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.00538166,\"nuts_mean\":-0.033749,\"nuts_sd\":0.187156,\"site\":\"s43\",\"svi_mean\":-0.0255505,\"svi_sd\":0.188183},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.539611,\"nuts_mean\":-0.129131,\"nuts_sd\":0.211082,\"site\":\"s44\",\"svi_mean\":-0.130591,\"svi_sd\":0.218494},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.151511,\"nuts_mean\":-0.11822,\"nuts_sd\":0.148758,\"site\":\"s45\",\"svi_mean\":-0.116223,\"svi_sd\":0.148462},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00596017,\"nuts_mean\":0.442741,\"nuts_sd\":0.167723,\"site\":\"s46\",\"svi_mean\":0.439963,\"svi_sd\":0.171516},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.169441,\"nuts_mean\":0.0260763,\"nuts_sd\":0.204366,\"site\":\"s47\",\"svi_mean\":0.0272432,\"svi_sd\":0.204161},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.273623,\"nuts_mean\":-0.122949,\"nuts_sd\":0.150173,\"site\":\"s48\",\"svi_mean\":-0.125767,\"svi_sd\":0.15283},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0733819,\"nuts_mean\":-0.302718,\"nuts_sd\":0.152408,\"site\":\"s49\",\"svi_mean\":-0.304106,\"svi_sd\":0.147842},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0771014,\"nuts_mean\":-0.29003,\"nuts_sd\":0.160136,\"site\":\"s50\",\"svi_mean\":-0.286453,\"svi_sd\":0.158972},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.155737,\"nuts_mean\":0.361389,\"nuts_sd\":0.12577,\"site\":\"s51\",\"svi_mean\":0.362067,\"svi_sd\":0.124754},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0336616,\"nuts_mean\":-0.00673754,\"nuts_sd\":0.208859,\"site\":\"s52\",\"svi_mean\":-0.00383735,\"svi_sd\":0.204779},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.361336,\"nuts_mean\":-0.121109,\"nuts_sd\":0.178712,\"site\":\"s53\",\"svi_mean\":-0.117989,\"svi_sd\":0.178829},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.447757,\"nuts_mean\":-0.242832,\"nuts_sd\":0.190979,\"site\":\"s54\",\"svi_mean\":-0.241716,\"svi_sd\":0.191875},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00311238,\"nuts_mean\":0.152418,\"nuts_sd\":0.148144,\"site\":\"s55\",\"svi_mean\":0.154219,\"svi_sd\":0.149965},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.148032,\"nuts_mean\":0.200859,\"nuts_sd\":0.16488,\"site\":\"s56\",\"svi_mean\":0.193862,\"svi_sd\":0.168076},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.568799,\"nuts_mean\":-0.28153,\"nuts_sd\":0.213554,\"site\":\"s57\",\"svi_mean\":-0.270973,\"svi_sd\":0.211041},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0641479,\"nuts_mean\":-0.071174,\"nuts_sd\":0.182687,\"site\":\"s58\",\"svi_mean\":-0.0734542,\"svi_sd\":0.182559},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0301053,\"nuts_mean\":0.146445,\"nuts_sd\":0.187395,\"site\":\"s59\",\"svi_mean\":0.142644,\"svi_sd\":0.188279},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.326973,\"nuts_mean\":0.3707,\"nuts_sd\":0.0804102,\"site\":\"s60\",\"svi_mean\":0.368136,\"svi_sd\":0.077895},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.368827,\"nuts_mean\":-0.0323173,\"nuts_sd\":0.154229,\"site\":\"s61\",\"svi_mean\":-0.0363928,\"svi_sd\":0.158126},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0432038,\"nuts_mean\":-0.398928,\"nuts_sd\":0.170059,\"site\":\"s62\",\"svi_mean\":-0.402364,\"svi_sd\":0.166034},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.215058,\"nuts_mean\":-0.125611,\"nuts_sd\":0.135088,\"site\":\"s63\",\"svi_mean\":-0.12327,\"svi_sd\":0.136595},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.15362,\"nuts_mean\":-0.281931,\"nuts_sd\":0.21638,\"site\":\"s64\",\"svi_mean\":-0.274652,\"svi_sd\":0.218608},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.590059,\"nuts_mean\":-0.296138,\"nuts_sd\":0.163991,\"site\":\"s65\",\"svi_mean\":-0.29121,\"svi_sd\":0.161851},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.18622,\"nuts_mean\":-0.0586629,\"nuts_sd\":0.190801,\"site\":\"s66\",\"svi_mean\":-0.0624576,\"svi_sd\":0.185533},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.30788,\"nuts_mean\":0.219326,\"nuts_sd\":0.168118,\"site\":\"s67\",\"svi_mean\":0.221061,\"svi_sd\":0.169635},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.127601,\"nuts_mean\":0.273287,\"nuts_sd\":0.150798,\"site\":\"s68\",\"svi_mean\":0.271724,\"svi_sd\":0.149861},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.393979,\"nuts_mean\":0.0918803,\"nuts_sd\":0.198228,\"site\":\"s69\",\"svi_mean\":0.0852639,\"svi_sd\":0.197506},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0153952,\"nuts_mean\":0.27492,\"nuts_sd\":0.160117,\"site\":\"s70\",\"svi_mean\":0.275307,\"svi_sd\":0.158094},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.180583,\"nuts_mean\":-0.0779803,\"nuts_sd\":0.212912,\"site\":\"s71\",\"svi_mean\":-0.0898729,\"svi_sd\":0.212611},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.113085,\"nuts_mean\":-0.000124026,\"nuts_sd\":0.148831,\"site\":\"s72\",\"svi_mean\":0.00071392,\"svi_sd\":0.148087},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0429174,\"nuts_mean\":0.0484821,\"nuts_sd\":0.174476,\"site\":\"s73\",\"svi_mean\":0.046625,\"svi_sd\":0.174096},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.162945,\"nuts_mean\":-0.476359,\"nuts_sd\":0.201699,\"site\":\"s74\",\"svi_mean\":-0.468608,\"svi_sd\":0.201661},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.114633,\"nuts_mean\":0.181454,\"nuts_sd\":0.153269,\"site\":\"s75\",\"svi_mean\":0.181752,\"svi_sd\":0.148553},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0920041,\"nuts_mean\":-0.102732,\"nuts_sd\":0.226326,\"site\":\"s76\",\"svi_mean\":-0.093375,\"svi_sd\":0.224322},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.218041,\"nuts_mean\":0.050327,\"nuts_sd\":0.125167,\"site\":\"s77\",\"svi_mean\":0.0525896,\"svi_sd\":0.12429},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0196956,\"nuts_mean\":-0.404763,\"nuts_sd\":0.128002,\"site\":\"s78\",\"svi_mean\":-0.405939,\"svi_sd\":0.123249},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0846015,\"nuts_mean\":0.367562,\"nuts_sd\":0.098832,\"site\":\"s79\",\"svi_mean\":0.366665,\"svi_sd\":0.0949674},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.654992,\"nuts_mean\":0.229506,\"nuts_sd\":0.110632,\"site\":\"s80\",\"svi_mean\":0.236847,\"svi_sd\":0.112609},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.110384,\"nuts_mean\":-0.264944,\"nuts_sd\":0.18047,\"site\":\"s81\",\"svi_mean\":-0.267141,\"svi_sd\":0.17912},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.174426,\"nuts_mean\":0.578612,\"nuts_sd\":0.105923,\"site\":\"s82\",\"svi_mean\":0.579518,\"svi_sd\":0.107028},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0347126,\"nuts_mean\":-0.000134949,\"nuts_sd\":0.201993,\"site\":\"s83\",\"svi_mean\":-0.00701798,\"svi_sd\":0.197578},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.247209,\"nuts_mean\":0.0781302,\"nuts_sd\":0.232034,\"site\":\"s84\",\"svi_mean\":0.080877,\"svi_sd\":0.223347},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.376882,\"nuts_mean\":-0.101329,\"nuts_sd\":0.201609,\"site\":\"s85\",\"svi_mean\":-0.0970229,\"svi_sd\":0.193553},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.17636,\"nuts_mean\":-0.0692846,\"nuts_sd\":0.174484,\"site\":\"s86\",\"svi_mean\":-0.0753278,\"svi_sd\":0.171986},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.263703,\"nuts_mean\":0.068532,\"nuts_sd\":0.160254,\"site\":\"s87\",\"svi_mean\":0.0720718,\"svi_sd\":0.160278},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.12768,\"nuts_mean\":-0.0587798,\"nuts_sd\":0.207476,\"site\":\"s88\",\"svi_mean\":-0.0642972,\"svi_sd\":0.214792},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0292716,\"nuts_mean\":-0.069936,\"nuts_sd\":0.204177,\"site\":\"s89\",\"svi_mean\":-0.0668576,\"svi_sd\":0.2016},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.217064,\"nuts_mean\":-0.0309486,\"nuts_sd\":0.175171,\"site\":\"s90\",\"svi_mean\":-0.0280903,\"svi_sd\":0.176381},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.257041,\"nuts_mean\":0.0819016,\"nuts_sd\":0.198943,\"site\":\"s91\",\"svi_mean\":0.082852,\"svi_sd\":0.197637},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0516233,\"nuts_mean\":0.116225,\"nuts_sd\":0.199319,\"site\":\"s92\",\"svi_mean\":0.112591,\"svi_sd\":0.197442},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.618958,\"nuts_mean\":0.0917515,\"nuts_sd\":0.209026,\"site\":\"s93\",\"svi_mean\":0.0848122,\"svi_sd\":0.21066},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.176827,\"nuts_mean\":-0.0102409,\"nuts_sd\":0.158859,\"site\":\"s94\",\"svi_mean\":-0.00077405,\"svi_sd\":0.155603},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0204884,\"nuts_mean\":-0.165335,\"nuts_sd\":0.222589,\"site\":\"s95\",\"svi_mean\":-0.162526,\"svi_sd\":0.222481},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.130979,\"nuts_mean\":0.173269,\"nuts_sd\":0.126378,\"site\":\"s96\",\"svi_mean\":0.17137,\"svi_sd\":0.125916},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.17802,\"nuts_mean\":0.254152,\"nuts_sd\":0.157022,\"site\":\"s97\",\"svi_mean\":0.253633,\"svi_sd\":0.156665},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.352621,\"nuts_mean\":0.0876908,\"nuts_sd\":0.159421,\"site\":\"s98\",\"svi_mean\":0.0843175,\"svi_sd\":0.154018},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.545751,\"nuts_mean\":-0.0718031,\"nuts_sd\":0.223057,\"site\":\"s99\",\"svi_mean\":-0.0699477,\"svi_sd\":0.222315},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.234866,\"nuts_mean\":0.0770934,\"nuts_sd\":0.169663,\"site\":\"s100\",\"svi_mean\":0.0774492,\"svi_sd\":0.172654},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.332977,\"nuts_mean\":-0.199014,\"nuts_sd\":0.172356,\"site\":\"s101\",\"svi_mean\":-0.198653,\"svi_sd\":0.170153},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.14222,\"nuts_mean\":0.222875,\"nuts_sd\":0.183727,\"site\":\"s102\",\"svi_mean\":0.221955,\"svi_sd\":0.17789},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.5583,\"nuts_mean\":-0.162818,\"nuts_sd\":0.218028,\"site\":\"s103\",\"svi_mean\":-0.164936,\"svi_sd\":0.221731},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0642757,\"nuts_mean\":0.0702509,\"nuts_sd\":0.151397,\"site\":\"s104\",\"svi_mean\":0.0759543,\"svi_sd\":0.144115},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.269922,\"nuts_mean\":-0.159955,\"nuts_sd\":0.145335,\"site\":\"s105\",\"svi_mean\":-0.163264,\"svi_sd\":0.139436},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.418185,\"nuts_mean\":-0.0814791,\"nuts_sd\":0.183547,\"site\":\"s106\",\"svi_mean\":-0.0771879,\"svi_sd\":0.185408},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0445377,\"nuts_mean\":0.323874,\"nuts_sd\":0.114033,\"site\":\"s107\",\"svi_mean\":0.322978,\"svi_sd\":0.11015},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.133709,\"nuts_mean\":0.0481967,\"nuts_sd\":0.162759,\"site\":\"s108\",\"svi_mean\":0.049188,\"svi_sd\":0.161408},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.130545,\"nuts_mean\":-0.00754294,\"nuts_sd\":0.218153,\"site\":\"s109\",\"svi_mean\":-0.00810619,\"svi_sd\":0.21156},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.410446,\"nuts_mean\":0.213077,\"nuts_sd\":0.179666,\"site\":\"s110\",\"svi_mean\":0.212393,\"svi_sd\":0.177481},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0186579,\"nuts_mean\":0.137716,\"nuts_sd\":0.14626,\"site\":\"s111\",\"svi_mean\":0.142282,\"svi_sd\":0.138942},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0543393,\"nuts_mean\":0.0755133,\"nuts_sd\":0.177554,\"site\":\"s112\",\"svi_mean\":0.0808282,\"svi_sd\":0.170025},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0251281,\"nuts_mean\":-0.155093,\"nuts_sd\":0.201575,\"site\":\"s113\",\"svi_mean\":-0.153272,\"svi_sd\":0.202901},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.585321,\"nuts_mean\":0.108818,\"nuts_sd\":0.21404,\"site\":\"s114\",\"svi_mean\":0.103879,\"svi_sd\":0.211829},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.206993,\"nuts_mean\":-0.0432172,\"nuts_sd\":0.130675,\"site\":\"s115\",\"svi_mean\":-0.0453841,\"svi_sd\":0.134414},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.155847,\"nuts_mean\":0.0879359,\"nuts_sd\":0.146303,\"site\":\"s116\",\"svi_mean\":0.0902471,\"svi_sd\":0.14653},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.6342,\"nuts_mean\":0.0315721,\"nuts_sd\":0.10754,\"site\":\"s117\",\"svi_mean\":0.0352527,\"svi_sd\":0.109308},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.266215,\"nuts_mean\":0.020787,\"nuts_sd\":0.224734,\"site\":\"s118\",\"svi_mean\":0.0210291,\"svi_sd\":0.234144},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0761817,\"nuts_mean\":0.083378,\"nuts_sd\":0.196258,\"site\":\"s119\",\"svi_mean\":0.0854362,\"svi_sd\":0.200333},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.136771,\"nuts_mean\":0.153929,\"nuts_sd\":0.174264,\"site\":\"s120\",\"svi_mean\":0.157226,\"svi_sd\":0.171478},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0701424,\"nuts_mean\":0.0531236,\"nuts_sd\":0.17161,\"site\":\"s121\",\"svi_mean\":0.0474405,\"svi_sd\":0.17309},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0420374,\"nuts_mean\":-0.28566,\"nuts_sd\":0.155941,\"site\":\"s122\",\"svi_mean\":-0.284495,\"svi_sd\":0.156847},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.042244,\"nuts_mean\":0.624096,\"nuts_sd\":0.115669,\"site\":\"s123\",\"svi_mean\":0.624927,\"svi_sd\":0.118741},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.134586,\"nuts_mean\":-0.0870591,\"nuts_sd\":0.155262,\"site\":\"s124\",\"svi_mean\":-0.0825797,\"svi_sd\":0.156046},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.158964,\"nuts_mean\":0.0653153,\"nuts_sd\":0.134052,\"site\":\"s125\",\"svi_mean\":0.0656792,\"svi_sd\":0.130674},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0140215,\"nuts_mean\":0.238365,\"nuts_sd\":0.188082,\"site\":\"s126\",\"svi_mean\":0.231599,\"svi_sd\":0.18856},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.112756,\"nuts_mean\":0.394283,\"nuts_sd\":0.109022,\"site\":\"s127\",\"svi_mean\":0.396684,\"svi_sd\":0.109777},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0345293,\"nuts_mean\":-0.07976,\"nuts_sd\":0.184938,\"site\":\"s128\",\"svi_mean\":-0.0765602,\"svi_sd\":0.181605},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0492122,\"nuts_mean\":0.146924,\"nuts_sd\":0.221657,\"site\":\"s129\",\"svi_mean\":0.146227,\"svi_sd\":0.218158},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.378458,\"nuts_mean\":0.308388,\"nuts_sd\":0.175365,\"site\":\"s130\",\"svi_mean\":0.309049,\"svi_sd\":0.175136},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.189653,\"nuts_mean\":-0.028147,\"nuts_sd\":0.148015,\"site\":\"s131\",\"svi_mean\":-0.0250648,\"svi_sd\":0.147567},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0768027,\"nuts_mean\":0.0266562,\"nuts_sd\":0.123678,\"site\":\"s132\",\"svi_mean\":0.0194772,\"svi_sd\":0.124834},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.146147,\"nuts_mean\":0.202629,\"nuts_sd\":0.112929,\"site\":\"s133\",\"svi_mean\":0.201949,\"svi_sd\":0.110845},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0746588,\"nuts_mean\":0.0316222,\"nuts_sd\":0.189654,\"site\":\"s134\",\"svi_mean\":0.037843,\"svi_sd\":0.188647},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.525896,\"nuts_mean\":-0.396083,\"nuts_sd\":0.17273,\"site\":\"s135\",\"svi_mean\":-0.387077,\"svi_sd\":0.172982},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0285995,\"nuts_mean\":0.14102,\"nuts_sd\":0.134856,\"site\":\"s136\",\"svi_mean\":0.143093,\"svi_sd\":0.13345},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0424894,\"nuts_mean\":-0.240605,\"nuts_sd\":0.144944,\"site\":\"s137\",\"svi_mean\":-0.241389,\"svi_sd\":0.143036},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.260528,\"nuts_mean\":-0.164369,\"nuts_sd\":0.161678,\"site\":\"s138\",\"svi_mean\":-0.167138,\"svi_sd\":0.162593},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0643486,\"nuts_mean\":0.0296237,\"nuts_sd\":0.156914,\"site\":\"s139\",\"svi_mean\":0.025179,\"svi_sd\":0.155115},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.12597,\"nuts_mean\":0.121446,\"nuts_sd\":0.173519,\"site\":\"s140\",\"svi_mean\":0.123542,\"svi_sd\":0.172585},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.300869,\"nuts_mean\":0.293289,\"nuts_sd\":0.198774,\"site\":\"s141\",\"svi_mean\":0.294514,\"svi_sd\":0.199492},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.13859,\"nuts_mean\":-0.0419474,\"nuts_sd\":0.169608,\"site\":\"s142\",\"svi_mean\":-0.0447777,\"svi_sd\":0.175692},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.43808,\"nuts_mean\":-0.0976846,\"nuts_sd\":0.166143,\"site\":\"s143\",\"svi_mean\":-0.10111,\"svi_sd\":0.167505},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.246954,\"nuts_mean\":0.287501,\"nuts_sd\":0.13708,\"site\":\"s144\",\"svi_mean\":0.288883,\"svi_sd\":0.134262},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0664493,\"nuts_mean\":0.168596,\"nuts_sd\":0.158649,\"site\":\"s145\",\"svi_mean\":0.167819,\"svi_sd\":0.156711},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0379634,\"nuts_mean\":-0.152548,\"nuts_sd\":0.166665,\"site\":\"s146\",\"svi_mean\":-0.147233,\"svi_sd\":0.159865},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.387619,\"nuts_mean\":-0.133898,\"nuts_sd\":0.209792,\"site\":\"s147\",\"svi_mean\":-0.131858,\"svi_sd\":0.207611},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0254897,\"nuts_mean\":-0.0467501,\"nuts_sd\":0.119439,\"site\":\"s148\",\"svi_mean\":-0.045494,\"svi_sd\":0.120869},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.579946,\"nuts_mean\":0.587607,\"nuts_sd\":0.0985731,\"site\":\"s149\",\"svi_mean\":0.582477,\"svi_sd\":0.0951805},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.217912,\"nuts_mean\":0.0400199,\"nuts_sd\":0.1385,\"site\":\"s150\",\"svi_mean\":0.0365273,\"svi_sd\":0.134167},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.190875,\"nuts_mean\":-0.315531,\"nuts_sd\":0.151552,\"site\":\"s151\",\"svi_mean\":-0.318553,\"svi_sd\":0.151903},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.339765,\"nuts_mean\":0.00372519,\"nuts_sd\":0.227781,\"site\":\"s152\",\"svi_mean\":0.00154538,\"svi_sd\":0.230552},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.317153,\"nuts_mean\":-0.181862,\"nuts_sd\":0.15093,\"site\":\"s153\",\"svi_mean\":-0.177623,\"svi_sd\":0.145378},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.425676,\"nuts_mean\":-0.223892,\"nuts_sd\":0.159798,\"site\":\"s154\",\"svi_mean\":-0.22075,\"svi_sd\":0.156576},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.12389,\"nuts_mean\":-0.204462,\"nuts_sd\":0.22953,\"site\":\"s155\",\"svi_mean\":-0.203602,\"svi_sd\":0.224875},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.106044,\"nuts_mean\":-0.155796,\"nuts_sd\":0.177336,\"site\":\"s156\",\"svi_mean\":-0.161946,\"svi_sd\":0.171889},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.443538,\"nuts_mean\":0.00246628,\"nuts_sd\":0.161485,\"site\":\"s157\",\"svi_mean\":-0.00855045,\"svi_sd\":0.162088},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.227579,\"nuts_mean\":-0.0610547,\"nuts_sd\":0.19282,\"site\":\"s158\",\"svi_mean\":-0.0636828,\"svi_sd\":0.194016},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.298343,\"nuts_mean\":0.339927,\"nuts_sd\":0.102458,\"site\":\"s159\",\"svi_mean\":0.340398,\"svi_sd\":0.101808},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.245774,\"nuts_mean\":-0.0725798,\"nuts_sd\":0.186518,\"site\":\"s160\",\"svi_mean\":-0.0713094,\"svi_sd\":0.187858},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.214319,\"nuts_mean\":0.0463823,\"nuts_sd\":0.208376,\"site\":\"s161\",\"svi_mean\":0.0571078,\"svi_sd\":0.209164},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.318519,\"nuts_mean\":0.0580749,\"nuts_sd\":0.158477,\"site\":\"s162\",\"svi_mean\":0.0524517,\"svi_sd\":0.156849},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.157474,\"nuts_mean\":-0.311867,\"nuts_sd\":0.178884,\"site\":\"s163\",\"svi_mean\":-0.307739,\"svi_sd\":0.179434},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0840705,\"nuts_mean\":-0.227671,\"nuts_sd\":0.199999,\"site\":\"s164\",\"svi_mean\":-0.229259,\"svi_sd\":0.199817},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.264831,\"nuts_mean\":0.0869354,\"nuts_sd\":0.206377,\"site\":\"s165\",\"svi_mean\":0.0916536,\"svi_sd\":0.20314},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0523511,\"nuts_mean\":-0.165956,\"nuts_sd\":0.121891,\"site\":\"s166\",\"svi_mean\":-0.164945,\"svi_sd\":0.121652},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.233131,\"nuts_mean\":-0.0218265,\"nuts_sd\":0.207906,\"site\":\"s167\",\"svi_mean\":-0.0185085,\"svi_sd\":0.207966},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00251025,\"nuts_mean\":-0.189549,\"nuts_sd\":0.148079,\"site\":\"s168\",\"svi_mean\":-0.184109,\"svi_sd\":0.145015},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.483383,\"nuts_mean\":-0.0781972,\"nuts_sd\":0.189239,\"site\":\"s169\",\"svi_mean\":-0.0753361,\"svi_sd\":0.186875},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.133094,\"nuts_mean\":-0.104677,\"nuts_sd\":0.185152,\"site\":\"s170\",\"svi_mean\":-0.0961963,\"svi_sd\":0.184201},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.657094,\"nuts_mean\":-0.218628,\"nuts_sd\":0.156714,\"site\":\"s171\",\"svi_mean\":-0.217708,\"svi_sd\":0.159613},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.914056,\"nuts_mean\":0.0476324,\"nuts_sd\":0.13313,\"site\":\"s172\",\"svi_mean\":0.0496004,\"svi_sd\":0.134621},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.275412,\"nuts_mean\":-0.124059,\"nuts_sd\":0.20083,\"site\":\"s173\",\"svi_mean\":-0.118307,\"svi_sd\":0.19797},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0467515,\"nuts_mean\":0.0243979,\"nuts_sd\":0.198973,\"site\":\"s174\",\"svi_mean\":0.0177221,\"svi_sd\":0.198772},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.133066,\"nuts_mean\":0.127266,\"nuts_sd\":0.198731,\"site\":\"s175\",\"svi_mean\":0.12657,\"svi_sd\":0.19235},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.328251,\"nuts_mean\":-0.0481289,\"nuts_sd\":0.16438,\"site\":\"s176\",\"svi_mean\":-0.0486142,\"svi_sd\":0.168847},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.168954,\"nuts_mean\":0.348675,\"nuts_sd\":0.132468,\"site\":\"s177\",\"svi_mean\":0.348973,\"svi_sd\":0.128518},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.28357,\"nuts_mean\":-0.12768,\"nuts_sd\":0.199714,\"site\":\"s178\",\"svi_mean\":-0.121087,\"svi_sd\":0.191823},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.133979,\"nuts_mean\":-0.196466,\"nuts_sd\":0.177715,\"site\":\"s179\",\"svi_mean\":-0.203292,\"svi_sd\":0.176957},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.066766,\"nuts_mean\":-0.10148,\"nuts_sd\":0.168065,\"site\":\"s180\",\"svi_mean\":-0.105226,\"svi_sd\":0.165801},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0769517,\"nuts_mean\":-0.0231884,\"nuts_sd\":0.161462,\"site\":\"s181\",\"svi_mean\":-0.0248962,\"svi_sd\":0.162128},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.229099,\"nuts_mean\":0.0722313,\"nuts_sd\":0.152606,\"site\":\"s182\",\"svi_mean\":0.0719715,\"svi_sd\":0.150813},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.06192,\"nuts_mean\":-0.134735,\"nuts_sd\":0.174504,\"site\":\"s183\",\"svi_mean\":-0.139804,\"svi_sd\":0.178572},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.29476,\"nuts_mean\":-0.0143642,\"nuts_sd\":0.149151,\"site\":\"s184\",\"svi_mean\":-0.0118296,\"svi_sd\":0.152584},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.159441,\"nuts_mean\":-0.0504191,\"nuts_sd\":0.167379,\"site\":\"s185\",\"svi_mean\":-0.0572118,\"svi_sd\":0.16675},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.195032,\"nuts_mean\":0.449181,\"nuts_sd\":0.0840789,\"site\":\"s186\",\"svi_mean\":0.447716,\"svi_sd\":0.084158},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.302629,\"nuts_mean\":-0.395205,\"nuts_sd\":0.163416,\"site\":\"s187\",\"svi_mean\":-0.383822,\"svi_sd\":0.167225},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.146612,\"nuts_mean\":-0.238645,\"nuts_sd\":0.163734,\"site\":\"s188\",\"svi_mean\":-0.234449,\"svi_sd\":0.162726},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.15791,\"nuts_mean\":-0.02261,\"nuts_sd\":0.18033,\"site\":\"s189\",\"svi_mean\":-0.0174335,\"svi_sd\":0.181167},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0112542,\"nuts_mean\":-0.0975661,\"nuts_sd\":0.208367,\"site\":\"s190\",\"svi_mean\":-0.0970673,\"svi_sd\":0.209936},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.297157,\"nuts_mean\":-0.30537,\"nuts_sd\":0.119523,\"site\":\"s191\",\"svi_mean\":-0.298903,\"svi_sd\":0.119676},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.220318,\"nuts_mean\":-0.173007,\"nuts_sd\":0.165885,\"site\":\"s192\",\"svi_mean\":-0.173913,\"svi_sd\":0.166892},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.216717,\"nuts_mean\":0.137297,\"nuts_sd\":0.184089,\"site\":\"s193\",\"svi_mean\":0.141597,\"svi_sd\":0.185897},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.027681,\"nuts_mean\":-0.410995,\"nuts_sd\":0.184266,\"site\":\"s194\",\"svi_mean\":-0.407018,\"svi_sd\":0.176051},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.435383,\"nuts_mean\":-0.21151,\"nuts_sd\":0.220585,\"site\":\"s195\",\"svi_mean\":-0.208161,\"svi_sd\":0.21757},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.082602,\"nuts_mean\":0.231006,\"nuts_sd\":0.121124,\"site\":\"s196\",\"svi_mean\":0.237122,\"svi_sd\":0.118224},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0520919,\"nuts_mean\":0.394836,\"nuts_sd\":0.0907629,\"site\":\"s197\",\"svi_mean\":0.397712,\"svi_sd\":0.0918974},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.252334,\"nuts_mean\":0.0104341,\"nuts_sd\":0.165116,\"site\":\"s198\",\"svi_mean\":0.0116912,\"svi_sd\":0.167684},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0390905,\"nuts_mean\":-0.0848354,\"nuts_sd\":0.146116,\"site\":\"s199\",\"svi_mean\":-0.0851409,\"svi_sd\":0.149698},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.240396,\"nuts_mean\":-0.0442517,\"nuts_sd\":0.154635,\"site\":\"s200\",\"svi_mean\":-0.0442252,\"svi_sd\":0.151498},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0573117,\"nuts_mean\":-0.214056,\"nuts_sd\":0.187811,\"site\":\"s201\",\"svi_mean\":-0.211368,\"svi_sd\":0.187617},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.269098,\"nuts_mean\":-0.108794,\"nuts_sd\":0.217627,\"site\":\"s202\",\"svi_mean\":-0.111267,\"svi_sd\":0.217577},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0987892,\"nuts_mean\":-0.0647915,\"nuts_sd\":0.139514,\"site\":\"s203\",\"svi_mean\":-0.067955,\"svi_sd\":0.135711},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0105189,\"nuts_mean\":-0.165933,\"nuts_sd\":0.179026,\"site\":\"s204\",\"svi_mean\":-0.158007,\"svi_sd\":0.18103},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0171369,\"nuts_mean\":-0.0675215,\"nuts_sd\":0.194562,\"site\":\"s205\",\"svi_mean\":-0.0701353,\"svi_sd\":0.200149},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.132522,\"nuts_mean\":0.121956,\"nuts_sd\":0.20996,\"site\":\"s206\",\"svi_mean\":0.115534,\"svi_sd\":0.208595},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.353855,\"nuts_mean\":-0.172752,\"nuts_sd\":0.179584,\"site\":\"s207\",\"svi_mean\":-0.174037,\"svi_sd\":0.177223},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.56806,\"nuts_mean\":-0.0184473,\"nuts_sd\":0.190931,\"site\":\"s208\",\"svi_mean\":-0.0177468,\"svi_sd\":0.184718},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.654771,\"nuts_mean\":0.32005,\"nuts_sd\":0.19164,\"site\":\"s209\",\"svi_mean\":0.315075,\"svi_sd\":0.19227},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.333478,\"nuts_mean\":0.304195,\"nuts_sd\":0.156135,\"site\":\"s210\",\"svi_mean\":0.304724,\"svi_sd\":0.156003},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.548568,\"nuts_mean\":0.0280772,\"nuts_sd\":0.190715,\"site\":\"s211\",\"svi_mean\":0.0271547,\"svi_sd\":0.193902},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.263394,\"nuts_mean\":0.301669,\"nuts_sd\":0.0913143,\"site\":\"s212\",\"svi_mean\":0.300927,\"svi_sd\":0.0909927},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.450598,\"nuts_mean\":-0.202279,\"nuts_sd\":0.137052,\"site\":\"s213\",\"svi_mean\":-0.198312,\"svi_sd\":0.137378},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.159922,\"nuts_mean\":-0.12765,\"nuts_sd\":0.18514,\"site\":\"s214\",\"svi_mean\":-0.127728,\"svi_sd\":0.183151},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.358036,\"nuts_mean\":0.397228,\"nuts_sd\":0.138418,\"site\":\"s215\",\"svi_mean\":0.400505,\"svi_sd\":0.138081},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0904351,\"nuts_mean\":0.179824,\"nuts_sd\":0.208402,\"site\":\"s216\",\"svi_mean\":0.185334,\"svi_sd\":0.200802},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0245059,\"nuts_mean\":0.154603,\"nuts_sd\":0.180131,\"site\":\"s217\",\"svi_mean\":0.159311,\"svi_sd\":0.181311},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0738176,\"nuts_mean\":0.316083,\"nuts_sd\":0.202896,\"site\":\"s218\",\"svi_mean\":0.310323,\"svi_sd\":0.198037},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.127766,\"nuts_mean\":0.01642,\"nuts_sd\":0.195136,\"site\":\"s219\",\"svi_mean\":0.00977809,\"svi_sd\":0.191527},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0560274,\"nuts_mean\":0.397179,\"nuts_sd\":0.123455,\"site\":\"s220\",\"svi_mean\":0.392127,\"svi_sd\":0.120723},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.000643691,\"nuts_mean\":0.189536,\"nuts_sd\":0.156021,\"site\":\"s221\",\"svi_mean\":0.189513,\"svi_sd\":0.15481},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.119383,\"nuts_mean\":0.0813369,\"nuts_sd\":0.172103,\"site\":\"s222\",\"svi_mean\":0.0777994,\"svi_sd\":0.169366},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.31948,\"nuts_mean\":0.191878,\"nuts_sd\":0.144213,\"site\":\"s223\",\"svi_mean\":0.193321,\"svi_sd\":0.14562},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.153871,\"nuts_mean\":-0.0908711,\"nuts_sd\":0.160629,\"site\":\"s224\",\"svi_mean\":-0.0926653,\"svi_sd\":0.156651},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.323618,\"nuts_mean\":-0.0998728,\"nuts_sd\":0.113611,\"site\":\"s225\",\"svi_mean\":-0.101595,\"svi_sd\":0.112533},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0278722,\"nuts_mean\":0.0746652,\"nuts_sd\":0.196844,\"site\":\"s226\",\"svi_mean\":0.0740105,\"svi_sd\":0.200833},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00272314,\"nuts_mean\":-0.210761,\"nuts_sd\":0.164914,\"site\":\"s227\",\"svi_mean\":-0.212804,\"svi_sd\":0.165901},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.519395,\"nuts_mean\":0.253393,\"nuts_sd\":0.191596,\"site\":\"s228\",\"svi_mean\":0.246535,\"svi_sd\":0.189308},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.00989022,\"nuts_mean\":-0.0241187,\"nuts_sd\":0.193898,\"site\":\"s229\",\"svi_mean\":-0.0242354,\"svi_sd\":0.193709},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.259047,\"nuts_mean\":0.0503335,\"nuts_sd\":0.210416,\"site\":\"s230\",\"svi_mean\":0.046722,\"svi_sd\":0.209722},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0932088,\"nuts_mean\":0.0229108,\"nuts_sd\":0.212247,\"site\":\"s231\",\"svi_mean\":0.0226146,\"svi_sd\":0.204951},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.689135,\"nuts_mean\":-0.158418,\"nuts_sd\":0.157119,\"site\":\"s232\",\"svi_mean\":-0.158369,\"svi_sd\":0.154952},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0386173,\"nuts_mean\":-0.251568,\"nuts_sd\":0.170801,\"site\":\"s233\",\"svi_mean\":-0.25264,\"svi_sd\":0.163825},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.152659,\"nuts_mean\":0.157614,\"nuts_sd\":0.161522,\"site\":\"s234\",\"svi_mean\":0.160811,\"svi_sd\":0.161778},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.776815,\"nuts_mean\":-0.386078,\"nuts_sd\":0.184613,\"site\":\"s235\",\"svi_mean\":-0.38357,\"svi_sd\":0.181645},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.37418,\"nuts_mean\":0.0270188,\"nuts_sd\":0.164838,\"site\":\"s236\",\"svi_mean\":0.0301038,\"svi_sd\":0.16577},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.30011,\"nuts_mean\":-0.170168,\"nuts_sd\":0.166159,\"site\":\"s237\",\"svi_mean\":-0.172213,\"svi_sd\":0.164671},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.226012,\"nuts_mean\":-0.182915,\"nuts_sd\":0.156183,\"site\":\"s238\",\"svi_mean\":-0.179003,\"svi_sd\":0.157311},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0339922,\"nuts_mean\":-0.0908667,\"nuts_sd\":0.190598,\"site\":\"s239\",\"svi_mean\":-0.0882549,\"svi_sd\":0.186247},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.288614,\"nuts_mean\":-0.168594,\"nuts_sd\":0.159761,\"site\":\"s240\",\"svi_mean\":-0.174599,\"svi_sd\":0.160137},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.444592,\"nuts_mean\":-0.0293905,\"nuts_sd\":0.23835,\"site\":\"s241\",\"svi_mean\":-0.0334323,\"svi_sd\":0.237003},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.154659,\"nuts_mean\":0.0500731,\"nuts_sd\":0.133522,\"site\":\"s242\",\"svi_mean\":0.0473835,\"svi_sd\":0.131868},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.244021,\"nuts_mean\":0.134232,\"nuts_sd\":0.155499,\"site\":\"s243\",\"svi_mean\":0.138039,\"svi_sd\":0.158966},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.229556,\"nuts_mean\":-0.0315747,\"nuts_sd\":0.185599,\"site\":\"s244\",\"svi_mean\":-0.0332902,\"svi_sd\":0.184658},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.250143,\"nuts_mean\":0.347134,\"nuts_sd\":0.128509,\"site\":\"s245\",\"svi_mean\":0.347531,\"svi_sd\":0.127284},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.337054,\"nuts_mean\":0.121297,\"nuts_sd\":0.150215,\"site\":\"s246\",\"svi_mean\":0.121298,\"svi_sd\":0.146087},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.425247,\"nuts_mean\":0.34188,\"nuts_sd\":0.138042,\"site\":\"s247\",\"svi_mean\":0.339969,\"svi_sd\":0.136431},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.137736,\"nuts_mean\":0.435515,\"nuts_sd\":0.170055,\"site\":\"s248\",\"svi_mean\":0.432598,\"svi_sd\":0.174991},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.1597,\"nuts_mean\":-0.266394,\"nuts_sd\":0.166624,\"site\":\"s249\",\"svi_mean\":-0.268236,\"svi_sd\":0.166117},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0337526,\"nuts_mean\":0.274319,\"nuts_sd\":0.19843,\"site\":\"s250\",\"svi_mean\":0.274617,\"svi_sd\":0.192704},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.137237,\"nuts_mean\":0.134984,\"nuts_sd\":0.128328,\"site\":\"s251\",\"svi_mean\":0.138623,\"svi_sd\":0.128065},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0953539,\"nuts_mean\":0.203012,\"nuts_sd\":0.14445,\"site\":\"s252\",\"svi_mean\":0.202635,\"svi_sd\":0.140294},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.177954,\"nuts_mean\":-0.00652745,\"nuts_sd\":0.212628,\"site\":\"s253\",\"svi_mean\":-0.00860629,\"svi_sd\":0.219563},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.267809,\"nuts_mean\":-0.169656,\"nuts_sd\":0.146527,\"site\":\"s254\",\"svi_mean\":-0.167466,\"svi_sd\":0.142069},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.168799,\"nuts_mean\":0.0612735,\"nuts_sd\":0.195647,\"site\":\"s255\",\"svi_mean\":0.0654542,\"svi_sd\":0.193275},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.111259,\"nuts_mean\":-0.298196,\"nuts_sd\":0.200782,\"site\":\"s256\",\"svi_mean\":-0.29559,\"svi_sd\":0.19953},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0559531,\"nuts_mean\":0.185112,\"nuts_sd\":0.208011,\"site\":\"s257\",\"svi_mean\":0.191002,\"svi_sd\":0.206414},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.44795,\"nuts_mean\":-0.0640391,\"nuts_sd\":0.233439,\"site\":\"s258\",\"svi_mean\":-0.0608354,\"svi_sd\":0.23228},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0114614,\"nuts_mean\":0.0895115,\"nuts_sd\":0.225142,\"site\":\"s259\",\"svi_mean\":0.0903536,\"svi_sd\":0.223946},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0177078,\"nuts_mean\":0.1609,\"nuts_sd\":0.172619,\"site\":\"s260\",\"svi_mean\":0.160343,\"svi_sd\":0.171532},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00972347,\"nuts_mean\":0.0265367,\"nuts_sd\":0.153844,\"site\":\"s261\",\"svi_mean\":0.0281377,\"svi_sd\":0.14913},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.218771,\"nuts_mean\":0.458899,\"nuts_sd\":0.167023,\"site\":\"s262\",\"svi_mean\":0.463411,\"svi_sd\":0.170193},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0480587,\"nuts_mean\":0.316837,\"nuts_sd\":0.138405,\"site\":\"s263\",\"svi_mean\":0.311998,\"svi_sd\":0.137831},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0806461,\"nuts_mean\":-0.0967854,\"nuts_sd\":0.19004,\"site\":\"s264\",\"svi_mean\":-0.0970307,\"svi_sd\":0.186769},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0766052,\"nuts_mean\":-0.408786,\"nuts_sd\":0.206025,\"site\":\"s265\",\"svi_mean\":-0.407145,\"svi_sd\":0.20217},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.491014,\"nuts_mean\":0.184217,\"nuts_sd\":0.209868,\"site\":\"s266\",\"svi_mean\":0.174077,\"svi_sd\":0.208456},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.312368,\"nuts_mean\":-0.118464,\"nuts_sd\":0.214679,\"site\":\"s267\",\"svi_mean\":-0.120017,\"svi_sd\":0.213754},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0700421,\"nuts_mean\":0.0749024,\"nuts_sd\":0.201507,\"site\":\"s268\",\"svi_mean\":0.0721897,\"svi_sd\":0.198982},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.181892,\"nuts_mean\":0.0648641,\"nuts_sd\":0.194789,\"site\":\"s269\",\"svi_mean\":0.0627693,\"svi_sd\":0.196166},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.256292,\"nuts_mean\":-0.163901,\"nuts_sd\":0.197385,\"site\":\"s270\",\"svi_mean\":-0.169167,\"svi_sd\":0.189278},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.345632,\"nuts_mean\":0.056583,\"nuts_sd\":0.182729,\"site\":\"s271\",\"svi_mean\":0.063832,\"svi_sd\":0.18043},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.276429,\"nuts_mean\":-0.0717207,\"nuts_sd\":0.123754,\"site\":\"s272\",\"svi_mean\":-0.0701734,\"svi_sd\":0.123714},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.296319,\"nuts_mean\":0.315975,\"nuts_sd\":0.224486,\"site\":\"s273\",\"svi_mean\":0.320154,\"svi_sd\":0.2192},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.141708,\"nuts_mean\":0.452605,\"nuts_sd\":0.145037,\"site\":\"s274\",\"svi_mean\":0.453105,\"svi_sd\":0.14088},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.283061,\"nuts_mean\":-0.187175,\"nuts_sd\":0.128491,\"site\":\"s275\",\"svi_mean\":-0.189208,\"svi_sd\":0.129227},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.296982,\"nuts_mean\":-0.141048,\"nuts_sd\":0.142152,\"site\":\"s276\",\"svi_mean\":-0.13653,\"svi_sd\":0.139077},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0768577,\"nuts_mean\":0.00770736,\"nuts_sd\":0.184616,\"site\":\"s277\",\"svi_mean\":-0.000641998,\"svi_sd\":0.184944},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0735675,\"nuts_mean\":0.252263,\"nuts_sd\":0.184884,\"site\":\"s278\",\"svi_mean\":0.249143,\"svi_sd\":0.185081},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0401003,\"nuts_mean\":0.138512,\"nuts_sd\":0.125954,\"site\":\"s279\",\"svi_mean\":0.139621,\"svi_sd\":0.122923},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.309104,\"nuts_mean\":0.0809315,\"nuts_sd\":0.165104,\"site\":\"s280\",\"svi_mean\":0.0798879,\"svi_sd\":0.167116},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0174638,\"nuts_mean\":-0.092699,\"nuts_sd\":0.221598,\"site\":\"s281\",\"svi_mean\":-0.091565,\"svi_sd\":0.222216},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.233065,\"nuts_mean\":-0.166115,\"nuts_sd\":0.230352,\"site\":\"s282\",\"svi_mean\":-0.170302,\"svi_sd\":0.234688},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.00949113,\"nuts_mean\":-0.222135,\"nuts_sd\":0.14848,\"site\":\"s283\",\"svi_mean\":-0.223221,\"svi_sd\":0.147813},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.640872,\"nuts_mean\":-0.199397,\"nuts_sd\":0.192603,\"site\":\"s284\",\"svi_mean\":-0.197739,\"svi_sd\":0.193432},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.251395,\"nuts_mean\":-0.120013,\"nuts_sd\":0.174523,\"site\":\"s285\",\"svi_mean\":-0.118826,\"svi_sd\":0.171687},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0615054,\"nuts_mean\":-0.00767423,\"nuts_sd\":0.128101,\"site\":\"s286\",\"svi_mean\":-0.00475531,\"svi_sd\":0.128654},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.454403,\"nuts_mean\":0.247327,\"nuts_sd\":0.126802,\"site\":\"s287\",\"svi_mean\":0.244977,\"svi_sd\":0.124036},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.299268,\"nuts_mean\":-0.166574,\"nuts_sd\":0.147695,\"site\":\"s288\",\"svi_mean\":-0.169791,\"svi_sd\":0.150217},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0163222,\"nuts_mean\":-0.231866,\"nuts_sd\":0.151474,\"site\":\"s289\",\"svi_mean\":-0.227213,\"svi_sd\":0.14734},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.113138,\"nuts_mean\":-0.016943,\"nuts_sd\":0.150677,\"site\":\"s290\",\"svi_mean\":-0.0138095,\"svi_sd\":0.148762},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0462922,\"nuts_mean\":0.0587684,\"nuts_sd\":0.185272,\"site\":\"s291\",\"svi_mean\":0.0560015,\"svi_sd\":0.191567},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.352584,\"nuts_mean\":-0.0243859,\"nuts_sd\":0.136459,\"site\":\"s292\",\"svi_mean\":-0.0289574,\"svi_sd\":0.132648},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.277593,\"nuts_mean\":0.355832,\"nuts_sd\":0.0797788,\"site\":\"s293\",\"svi_mean\":0.354316,\"svi_sd\":0.0781599},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.464766,\"nuts_mean\":-0.0719793,\"nuts_sd\":0.17274,\"site\":\"s294\",\"svi_mean\":-0.0717514,\"svi_sd\":0.176207},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.430539,\"nuts_mean\":0.00485029,\"nuts_sd\":0.100861,\"site\":\"s295\",\"svi_mean\":0.00883161,\"svi_sd\":0.0989452},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.358975,\"nuts_mean\":-0.217101,\"nuts_sd\":0.209426,\"site\":\"s296\",\"svi_mean\":-0.207966,\"svi_sd\":0.210043},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.472023,\"nuts_mean\":-0.00689969,\"nuts_sd\":0.206051,\"site\":\"s297\",\"svi_mean\":-0.00679431,\"svi_sd\":0.206873},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.262579,\"nuts_mean\":-0.0437981,\"nuts_sd\":0.128354,\"site\":\"s298\",\"svi_mean\":-0.0461996,\"svi_sd\":0.126423},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0996922,\"nuts_mean\":-0.346603,\"nuts_sd\":0.156837,\"site\":\"s299\",\"svi_mean\":-0.340689,\"svi_sd\":0.152892},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0800156,\"nuts_mean\":-0.139503,\"nuts_sd\":0.167527,\"site\":\"s300\",\"svi_mean\":-0.141902,\"svi_sd\":0.162885},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.222359,\"nuts_mean\":0.0644998,\"nuts_sd\":0.175507,\"site\":\"s301\",\"svi_mean\":0.0625267,\"svi_sd\":0.178623},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.106367,\"nuts_mean\":-0.233099,\"nuts_sd\":0.21475,\"site\":\"s302\",\"svi_mean\":-0.231329,\"svi_sd\":0.212012},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.289013,\"nuts_mean\":-0.190701,\"nuts_sd\":0.179684,\"site\":\"s303\",\"svi_mean\":-0.188989,\"svi_sd\":0.175897},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.35382,\"nuts_mean\":-0.25189,\"nuts_sd\":0.152679,\"site\":\"s304\",\"svi_mean\":-0.247679,\"svi_sd\":0.14908},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.361472,\"nuts_mean\":0.201038,\"nuts_sd\":0.10524,\"site\":\"s305\",\"svi_mean\":0.202305,\"svi_sd\":0.104104},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.277452,\"nuts_mean\":0.049734,\"nuts_sd\":0.207752,\"site\":\"s306\",\"svi_mean\":0.0447025,\"svi_sd\":0.207794},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0155323,\"nuts_mean\":0.047767,\"nuts_sd\":0.262181,\"site\":\"s307\",\"svi_mean\":0.0471756,\"svi_sd\":0.246222},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.186089,\"nuts_mean\":0.0493973,\"nuts_sd\":0.176474,\"site\":\"s308\",\"svi_mean\":0.0485651,\"svi_sd\":0.180707},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.249868,\"nuts_mean\":0.0919062,\"nuts_sd\":0.187402,\"site\":\"s309\",\"svi_mean\":0.0853717,\"svi_sd\":0.187161},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0532897,\"nuts_mean\":-0.087194,\"nuts_sd\":0.201572,\"site\":\"s310\",\"svi_mean\":-0.0885179,\"svi_sd\":0.201461},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.1693,\"nuts_mean\":0.0113352,\"nuts_sd\":0.202216,\"site\":\"s311\",\"svi_mean\":0.00721553,\"svi_sd\":0.204654},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.132748,\"nuts_mean\":0.174408,\"nuts_sd\":0.134724,\"site\":\"s312\",\"svi_mean\":0.182414,\"svi_sd\":0.133331},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.014735,\"nuts_mean\":0.213864,\"nuts_sd\":0.155525,\"site\":\"s313\",\"svi_mean\":0.215652,\"svi_sd\":0.152156},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.374768,\"nuts_mean\":0.201731,\"nuts_sd\":0.131578,\"site\":\"s314\",\"svi_mean\":0.197883,\"svi_sd\":0.131867},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0984672,\"nuts_mean\":-0.0346089,\"nuts_sd\":0.171537,\"site\":\"s315\",\"svi_mean\":-0.0365031,\"svi_sd\":0.171448},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0563705,\"nuts_mean\":0.11365,\"nuts_sd\":0.180974,\"site\":\"s316\",\"svi_mean\":0.105033,\"svi_sd\":0.179257},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.119157,\"nuts_mean\":-0.213745,\"nuts_sd\":0.182298,\"site\":\"s317\",\"svi_mean\":-0.20938,\"svi_sd\":0.179642},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0437512,\"nuts_mean\":0.0576996,\"nuts_sd\":0.158965,\"site\":\"s318\",\"svi_mean\":0.0562912,\"svi_sd\":0.15914},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.304255,\"nuts_mean\":0.14782,\"nuts_sd\":0.222939,\"site\":\"s319\",\"svi_mean\":0.14616,\"svi_sd\":0.217887},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.503035,\"nuts_mean\":0.189908,\"nuts_sd\":0.179981,\"site\":\"s320\",\"svi_mean\":0.187927,\"svi_sd\":0.181306},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.396637,\"nuts_mean\":-0.00504179,\"nuts_sd\":0.118935,\"site\":\"s321\",\"svi_mean\":-0.00786893,\"svi_sd\":0.120533},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.116218,\"nuts_mean\":0.435673,\"nuts_sd\":0.0819918,\"site\":\"s322\",\"svi_mean\":0.436462,\"svi_sd\":0.0803488},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.671708,\"nuts_mean\":0.0623147,\"nuts_sd\":0.170871,\"site\":\"s323\",\"svi_mean\":0.0610771,\"svi_sd\":0.166852},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.172624,\"nuts_mean\":0.0719845,\"nuts_sd\":0.234215,\"site\":\"s324\",\"svi_mean\":0.0754976,\"svi_sd\":0.230619},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.13071,\"nuts_mean\":0.0776538,\"nuts_sd\":0.186511,\"site\":\"s325\",\"svi_mean\":0.0759065,\"svi_sd\":0.183437},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0159584,\"nuts_mean\":-0.218746,\"nuts_sd\":0.198717,\"site\":\"s326\",\"svi_mean\":-0.211005,\"svi_sd\":0.199165},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0902288,\"nuts_mean\":0.220097,\"nuts_sd\":0.122084,\"site\":\"s327\",\"svi_mean\":0.225844,\"svi_sd\":0.119403},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.115398,\"nuts_mean\":-0.212778,\"nuts_sd\":0.155008,\"site\":\"s328\",\"svi_mean\":-0.212986,\"svi_sd\":0.152143},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.285343,\"nuts_mean\":-0.149255,\"nuts_sd\":0.152115,\"site\":\"s329\",\"svi_mean\":-0.154231,\"svi_sd\":0.149276},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.193693,\"nuts_mean\":0.0464739,\"nuts_sd\":0.160737,\"site\":\"s330\",\"svi_mean\":0.0544429,\"svi_sd\":0.155036},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.236154,\"nuts_mean\":0.16618,\"nuts_sd\":0.116131,\"site\":\"s331\",\"svi_mean\":0.165839,\"svi_sd\":0.119344},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.105888,\"nuts_mean\":-0.0175059,\"nuts_sd\":0.183546,\"site\":\"s332\",\"svi_mean\":-0.0214401,\"svi_sd\":0.1762},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.05667,\"nuts_mean\":0.0383462,\"nuts_sd\":0.169454,\"site\":\"s333\",\"svi_mean\":0.0382323,\"svi_sd\":0.170819},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.424193,\"nuts_mean\":0.0195313,\"nuts_sd\":0.135471,\"site\":\"s334\",\"svi_mean\":0.0205837,\"svi_sd\":0.136183},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.270186,\"nuts_mean\":0.0381908,\"nuts_sd\":0.192092,\"site\":\"s335\",\"svi_mean\":0.0388346,\"svi_sd\":0.186903},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.233207,\"nuts_mean\":-0.424019,\"nuts_sd\":0.175229,\"site\":\"s336\",\"svi_mean\":-0.424314,\"svi_sd\":0.166744},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.415207,\"nuts_mean\":-0.189861,\"nuts_sd\":0.166857,\"site\":\"s337\",\"svi_mean\":-0.195324,\"svi_sd\":0.169569},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.393081,\"nuts_mean\":-0.196416,\"nuts_sd\":0.197865,\"site\":\"s338\",\"svi_mean\":-0.191108,\"svi_sd\":0.190682},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.415253,\"nuts_mean\":0.0710592,\"nuts_sd\":0.223559,\"site\":\"s339\",\"svi_mean\":0.0710046,\"svi_sd\":0.22174},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.117807,\"nuts_mean\":-0.39455,\"nuts_sd\":0.180202,\"site\":\"s340\",\"svi_mean\":-0.390542,\"svi_sd\":0.177001},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.137236,\"nuts_mean\":-0.0708686,\"nuts_sd\":0.114302,\"site\":\"s341\",\"svi_mean\":-0.0683133,\"svi_sd\":0.113076},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0904772,\"nuts_mean\":-0.0533228,\"nuts_sd\":0.16122,\"site\":\"s342\",\"svi_mean\":-0.054934,\"svi_sd\":0.16041},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0536656,\"nuts_mean\":0.0383208,\"nuts_sd\":0.159561,\"site\":\"s343\",\"svi_mean\":0.035826,\"svi_sd\":0.159952},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.233903,\"nuts_mean\":-0.163866,\"nuts_sd\":0.148614,\"site\":\"s344\",\"svi_mean\":-0.165859,\"svi_sd\":0.142929},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.661101,\"nuts_mean\":0.0365642,\"nuts_sd\":0.136555,\"site\":\"s345\",\"svi_mean\":0.0338513,\"svi_sd\":0.139338},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.141744,\"nuts_mean\":-0.0908259,\"nuts_sd\":0.167445,\"site\":\"s346\",\"svi_mean\":-0.0882318,\"svi_sd\":0.165795},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0829261,\"nuts_mean\":-0.145592,\"nuts_sd\":0.124137,\"site\":\"s347\",\"svi_mean\":-0.143633,\"svi_sd\":0.12238},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.091715,\"nuts_mean\":-0.0381047,\"nuts_sd\":0.142802,\"site\":\"s348\",\"svi_mean\":-0.0377062,\"svi_sd\":0.141609},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.264094,\"nuts_mean\":-0.214948,\"nuts_sd\":0.176574,\"site\":\"s349\",\"svi_mean\":-0.209888,\"svi_sd\":0.182998},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00487999,\"nuts_mean\":-0.41348,\"nuts_sd\":0.162219,\"site\":\"s350\",\"svi_mean\":-0.414366,\"svi_sd\":0.157186},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.104556,\"nuts_mean\":0.386539,\"nuts_sd\":0.114112,\"site\":\"s351\",\"svi_mean\":0.385273,\"svi_sd\":0.11412},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.21633,\"nuts_mean\":-0.105931,\"nuts_sd\":0.184505,\"site\":\"s352\",\"svi_mean\":-0.106178,\"svi_sd\":0.186757},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.1075,\"nuts_mean\":0.0746356,\"nuts_sd\":0.163451,\"site\":\"s353\",\"svi_mean\":0.0705045,\"svi_sd\":0.160502},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0672754,\"nuts_mean\":0.322298,\"nuts_sd\":0.127667,\"site\":\"s354\",\"svi_mean\":0.320706,\"svi_sd\":0.123984},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0937212,\"nuts_mean\":-0.140763,\"nuts_sd\":0.200373,\"site\":\"s355\",\"svi_mean\":-0.140418,\"svi_sd\":0.192579},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0588334,\"nuts_mean\":-0.0508703,\"nuts_sd\":0.129506,\"site\":\"s356\",\"svi_mean\":-0.0495336,\"svi_sd\":0.128955},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.556353,\"nuts_mean\":-0.0715204,\"nuts_sd\":0.184449,\"site\":\"s357\",\"svi_mean\":-0.072529,\"svi_sd\":0.189499},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.460266,\"nuts_mean\":0.214435,\"nuts_sd\":0.160954,\"site\":\"s358\",\"svi_mean\":0.211409,\"svi_sd\":0.163014},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.53913,\"nuts_mean\":-0.312206,\"nuts_sd\":0.14081,\"site\":\"s359\",\"svi_mean\":-0.309854,\"svi_sd\":0.13974},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.164726,\"nuts_mean\":0.0735801,\"nuts_sd\":0.111745,\"site\":\"s360\",\"svi_mean\":0.0730386,\"svi_sd\":0.111025},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.105681,\"nuts_mean\":0.0876033,\"nuts_sd\":0.18153,\"site\":\"s361\",\"svi_mean\":0.0857201,\"svi_sd\":0.184681},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.394987,\"nuts_mean\":-0.0072146,\"nuts_sd\":0.17804,\"site\":\"s362\",\"svi_mean\":-0.0113768,\"svi_sd\":0.182718},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.112798,\"nuts_mean\":-0.137533,\"nuts_sd\":0.158861,\"site\":\"s363\",\"svi_mean\":-0.14235,\"svi_sd\":0.16275},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0578203,\"nuts_mean\":-0.184842,\"nuts_sd\":0.155835,\"site\":\"s364\",\"svi_mean\":-0.18463,\"svi_sd\":0.15323},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.024903,\"nuts_mean\":0.117728,\"nuts_sd\":0.211199,\"site\":\"s365\",\"svi_mean\":0.116419,\"svi_sd\":0.209482},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.1759,\"nuts_mean\":-0.0655924,\"nuts_sd\":0.187327,\"site\":\"s366\",\"svi_mean\":-0.0648334,\"svi_sd\":0.190619},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0505331,\"nuts_mean\":0.222727,\"nuts_sd\":0.152847,\"site\":\"s367\",\"svi_mean\":0.224238,\"svi_sd\":0.152381},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.146923,\"nuts_mean\":0.13687,\"nuts_sd\":0.124301,\"site\":\"s368\",\"svi_mean\":0.141736,\"svi_sd\":0.126758},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.306706,\"nuts_mean\":-0.0945441,\"nuts_sd\":0.104061,\"site\":\"s369\",\"svi_mean\":-0.100133,\"svi_sd\":0.105958},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.413907,\"nuts_mean\":0.21154,\"nuts_sd\":0.138985,\"site\":\"s370\",\"svi_mean\":0.204528,\"svi_sd\":0.133446},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.186258,\"nuts_mean\":0.00156407,\"nuts_sd\":0.178614,\"site\":\"s371\",\"svi_mean\":0.00264199,\"svi_sd\":0.179635},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.189014,\"nuts_mean\":0.0822992,\"nuts_sd\":0.227342,\"site\":\"s372\",\"svi_mean\":0.0804529,\"svi_sd\":0.228828},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.131586,\"nuts_mean\":0.067701,\"nuts_sd\":0.171299,\"site\":\"s373\",\"svi_mean\":0.0720881,\"svi_sd\":0.175956},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0169533,\"nuts_mean\":0.0263111,\"nuts_sd\":0.143657,\"site\":\"s374\",\"svi_mean\":0.0231759,\"svi_sd\":0.144636},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.11587,\"nuts_mean\":-0.124707,\"nuts_sd\":0.205055,\"site\":\"s375\",\"svi_mean\":-0.127343,\"svi_sd\":0.197133},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.27819,\"nuts_mean\":-0.0185153,\"nuts_sd\":0.16661,\"site\":\"s376\",\"svi_mean\":-0.0230116,\"svi_sd\":0.170247},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.128221,\"nuts_mean\":-0.0667593,\"nuts_sd\":0.129446,\"site\":\"s377\",\"svi_mean\":-0.0655737,\"svi_sd\":0.131072},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.00121079,\"nuts_mean\":-0.0561324,\"nuts_sd\":0.21844,\"site\":\"s378\",\"svi_mean\":-0.0670962,\"svi_sd\":0.21467},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.334951,\"nuts_mean\":0.372257,\"nuts_sd\":0.125808,\"site\":\"s379\",\"svi_mean\":0.36961,\"svi_sd\":0.129494},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.323645,\"nuts_mean\":0.490015,\"nuts_sd\":0.129372,\"site\":\"s380\",\"svi_mean\":0.486308,\"svi_sd\":0.127455},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.119208,\"nuts_mean\":-0.0137787,\"nuts_sd\":0.190789,\"site\":\"s381\",\"svi_mean\":-0.0140126,\"svi_sd\":0.192358},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0443104,\"nuts_mean\":0.208032,\"nuts_sd\":0.12715,\"site\":\"s382\",\"svi_mean\":0.207495,\"svi_sd\":0.124622},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.199308,\"nuts_mean\":-0.0655871,\"nuts_sd\":0.207208,\"site\":\"s383\",\"svi_mean\":-0.0618018,\"svi_sd\":0.199904},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.249699,\"nuts_mean\":0.21056,\"nuts_sd\":0.110299,\"site\":\"s384\",\"svi_mean\":0.212063,\"svi_sd\":0.107834},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.130553,\"nuts_mean\":-0.54774,\"nuts_sd\":0.121796,\"site\":\"s385\",\"svi_mean\":-0.552598,\"svi_sd\":0.123426},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.26983,\"nuts_mean\":0.223986,\"nuts_sd\":0.117879,\"site\":\"s386\",\"svi_mean\":0.224396,\"svi_sd\":0.117585},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.348089,\"nuts_mean\":0.0135619,\"nuts_sd\":0.144013,\"site\":\"s387\",\"svi_mean\":0.0102984,\"svi_sd\":0.148765},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.200515,\"nuts_mean\":-0.156376,\"nuts_sd\":0.143059,\"site\":\"s388\",\"svi_mean\":-0.157982,\"svi_sd\":0.142059},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.478752,\"nuts_mean\":0.0554093,\"nuts_sd\":0.129446,\"site\":\"s389\",\"svi_mean\":0.0575123,\"svi_sd\":0.12949},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0320101,\"nuts_mean\":0.0156609,\"nuts_sd\":0.144105,\"site\":\"s390\",\"svi_mean\":0.0158113,\"svi_sd\":0.138999},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.146304,\"nuts_mean\":-0.0635811,\"nuts_sd\":0.186999,\"site\":\"s391\",\"svi_mean\":-0.0616154,\"svi_sd\":0.193659},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.160602,\"nuts_mean\":0.0537547,\"nuts_sd\":0.173154,\"site\":\"s392\",\"svi_mean\":0.0570924,\"svi_sd\":0.179526},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.338313,\"nuts_mean\":-0.084924,\"nuts_sd\":0.192425,\"site\":\"s393\",\"svi_mean\":-0.0866344,\"svi_sd\":0.193563},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0718873,\"nuts_mean\":0.381206,\"nuts_sd\":0.116369,\"site\":\"s394\",\"svi_mean\":0.380852,\"svi_sd\":0.116536},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.276563,\"nuts_mean\":0.112158,\"nuts_sd\":0.180953,\"site\":\"s395\",\"svi_mean\":0.112208,\"svi_sd\":0.179868},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.0878737,\"nuts_mean\":-0.274252,\"nuts_sd\":0.147001,\"site\":\"s396\",\"svi_mean\":-0.271668,\"svi_sd\":0.143959},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.356399,\"nuts_mean\":0.0349961,\"nuts_sd\":0.14352,\"site\":\"s397\",\"svi_mean\":0.0377761,\"svi_sd\":0.139635},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.151089,\"nuts_mean\":-0.135107,\"nuts_sd\":0.122185,\"site\":\"s398\",\"svi_mean\":-0.132512,\"svi_sd\":0.124594},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":0.0831562,\"nuts_mean\":0.271935,\"nuts_sd\":0.0914752,\"site\":\"s399\",\"svi_mean\":0.272805,\"svi_sd\":0.090522},{\"__src\":\"_s1\",\"effect\":\"site slope deviation\",\"generating\":-0.668654,\"nuts_mean\":-0.278246,\"nuts_sd\":0.20194,\"site\":\"s400\",\"svi_mean\":-0.271595,\"svi_sd\":0.202429},{\"__src\":\"_s2\",\"effect\":\"site intercept deviation\",\"x\":-1.11016,\"y\":-1.11016},{\"__src\":\"_s2\",\"effect\":\"site slope deviation\",\"x\":-1.11016,\"y\":-1.11016},{\"__src\":\"_s2\",\"effect\":\"site intercept deviation\",\"x\":1.53177,\"y\":1.53177},{\"__src\":\"_s2\",\"effect\":\"site slope deviation\",\"x\":1.53177,\"y\":1.53177}]},\"facet\":{\"field\":\"effect\",\"type\":\"nominal\"},\"spec\":{\"height\":240,\"layer\":[{\"encoding\":{\"tooltip\":[{\"field\":\"nuts_mean\",\"type\":\"quantitative\"},{\"field\":\"svi_mean\",\"type\":\"quantitative\"},{\"field\":\"effect\",\"type\":\"nominal\"}],\"x\":{\"field\":\"nuts_mean\",\"scale\":{\"zero\":false},\"title\":\"NUTS posterior mean\",\"type\":\"quantitative\"},\"y\":{\"field\":\"svi_mean\",\"scale\":{\"zero\":false},\"title\":\"SVI posterior mean\",\"type\":\"quantitative\"}},\"mark\":{\"color\":\"#eb6834\",\"filled\":true,\"opacity\":0.6,\"size\":12.5,\"type\":\"point\"},\"params\":[{\"bind\":\"scales\",\"name\":\"grid\",\"select\":{\"encodings\":[\"x\",\"y\"],\"type\":\"interval\"}}],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},{\"encoding\":{\"tooltip\":[{\"field\":\"x\",\"type\":\"quantitative\"},{\"field\":\"y\",\"type\":\"quantitative\"},{\"field\":\"effect\",\"type\":\"nominal\"}],\"x\":{\"field\":\"x\",\"scale\":{\"zero\":false},\"title\":\"NUTS posterior mean\",\"type\":\"quantitative\"},\"y\":{\"field\":\"y\",\"scale\":{\"zero\":false},\"title\":\"SVI posterior mean\",\"type\":\"quantitative\"}},\"mark\":{\"strokeDash\":[6,6],\"style\":\"aov-ink\",\"type\":\"line\"},\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]}],\"width\":240},\"title\":\"Site effects\",\"description\":\"Each point is one site's intercept or slope deviation. Posterior means lie close to the equality line; the table also compares their uncertainty.\"}","caption":"Each point is one site's intercept or slope deviation. Posterior means lie close to the equality line; the table also compares their uncertainty.","alt":"Each point is one site's intercept or slope deviation. Posterior means lie close to the equality line; the table also compares their uncertainty."}| coordinates | count | mean error, median (NUTS sds) | mean error, largest | sd ratio SVI / NUTS: 5 %, median, 95 % |
|---|---|---|---|---|
| population parameters | 5 | 0.068 | 0.36 | 0.34, 0.39, 0.45 |
| site intercepts | 400 | 0.019 | 0.097 | 0.96, 0.99, 1 |
| site slopes | 400 | 0.016 | 0.072 | 0.96, 0.99, 1 |
Continue the fit as data arrive
A fitted approximation is a data-free summary of what was seen. brm_svi_checkpoint(fit) returns it as a BRMSVIState: the global factor, one factor per unit keyed by the unit's label, and the frozen preprocessing of the model. It holds no observations and no compiled model, so it serializes and travels. brm_svi_fit(state, builder(window)) then fits a new window of observations with that state as the prior and the start: the target is q_previous(globals, units) * p(window | globals, units).
Units are matched by label. A site seen before continues from its carried factor; a site that is new in a window starts from its authored prior under the carried global factor; a site absent from a window keeps its factor in the state and recovers it when it returns. Each observation enters once: the history is represented by the carried Gaussians and never revisited, so a stream of windows is a sequential approximation that can depend on the order of the windows and need not equal a fit of everything at once.
The stream here cuts the eight visits into four windows of two. The windows contain 300, 340, 380 and 400 sites. A site contributes observations only after it arrives; its earlier visits are absent from the accumulated reference data too:
# Window `w` holds visits `2w - 1` and `2w` of the sites present in it. Sites arrive
# over time: 300 in the first window, then 340, 380 and 400.
function stream_windows(data; n_windows=4, first_sites=300, new_sites_per_window=40)
visits = length(first(data.x))
per_window = visits ÷ n_windows
map(1:n_windows) do w
n = min(length(data.site), first_sites + (w - 1) * new_sites_per_window)
range = ((w - 1) * per_window + 1):(w * per_window)
(; site=data.site[1:n], x=[xi[range] for xi in data.x[1:n]], y=[yi[range] for yi in data.y[1:n]])
end
end# The stream: calibrate on the first window, then continue through the others,
# carrying only the checkpoint from one window to the next.
function stream_fits(windows; progress=nothing)
fit = brm_svi_fit(site_counts_model(windows[1]); batch_size=32, rng=Xoshiro(1), progress)
fits = Any[fit] # a continuation carries a BRMSVIState where the initial fit has nothing
for w in 2:length(windows)
state = brm_svi_checkpoint(fit)
fit = brm_svi_fit(state, site_counts_model(windows[w]); batch_size=32, rng=Xoshiro(w), progress)
push!(fits, fit)
end
fits
endAfter each window, the streaming estimate of every population parameter is compared with two fits of all the data seen up to that window: NUTS and a batch brm_svi_fit.
{"spec":"{\"$schema\":\"https://vega.github.io/schema/vega-lite/v5.json\",\"_aov\":{\"legendInteraction\":\"highlight\"},\"_aovSummary\":{\"bands\":[[\"q025\",\"q975\",\"q025/q975\"],[\"q25\",\"q75\",\"q25/q75\"]],\"point_col\":\"median\",\"value_label\":\"median\"},\"columns\":2,\"config\":{\"axis\":{\"labelLimit\":0},\"header\":{\"labelLimit\":0},\"legend\":{\"labelLimit\":0},\"style\":{\"aov-ink\":{\"color\":\"#333\"}}},\"data\":{\"values\":[{\"__src\":\"_s1\",\"mean\":1.06282,\"median\":1.0633,\"method\":\"NUTS\",\"parameter\":\"intercept a\",\"q025\":0.985202,\"q25\":1.03665,\"q75\":1.08932,\"q975\":1.13845,\"sd\":0.0391615,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":1.06189,\"median\":1.062,\"method\":\"streaming SVI\",\"parameter\":\"intercept a\",\"q025\":1.01779,\"q25\":1.04798,\"q75\":1.07658,\"q975\":1.10543,\"sd\":0.0219872,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.363373,\"median\":0.363416,\"method\":\"NUTS\",\"parameter\":\"slope b\",\"q025\":0.296389,\"q25\":0.34042,\"q75\":0.386701,\"q975\":0.429662,\"sd\":0.0340567,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.363038,\"median\":0.362994,\"method\":\"streaming SVI\",\"parameter\":\"slope b\",\"q025\":0.325118,\"q25\":0.349829,\"q75\":0.376511,\"q975\":0.400351,\"sd\":0.0193876,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.471317,\"median\":0.470567,\"method\":\"NUTS\",\"parameter\":\"sd of site intercepts\",\"q025\":0.401985,\"q25\":0.445845,\"q75\":0.495621,\"q975\":0.545271,\"sd\":0.0371485,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.466066,\"median\":0.465614,\"method\":\"streaming SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.429773,\"q25\":0.45353,\"q75\":0.477839,\"q975\":0.503879,\"sd\":0.018545,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.230142,\"median\":0.228933,\"method\":\"NUTS\",\"parameter\":\"sd of site slopes\",\"q025\":0.151283,\"q25\":0.200592,\"q75\":0.25792,\"q975\":0.314852,\"sd\":0.0423003,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.21351,\"median\":0.213005,\"method\":\"streaming SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.179565,\"q25\":0.202076,\"q75\":0.224175,\"q975\":0.249432,\"sd\":0.0175055,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.700886,\"median\":0.71209,\"method\":\"NUTS\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.343171,\"q25\":0.590275,\"q75\":0.829053,\"q975\":0.977941,\"sd\":0.16716,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":0.712446,\"median\":0.720105,\"method\":\"streaming SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.574916,\"q25\":0.67388,\"q75\":0.755691,\"q975\":0.815479,\"sd\":0.063013,\"window\":\"window 1\"},{\"__src\":\"_s1\",\"mean\":1.01769,\"median\":1.01802,\"method\":\"NUTS\",\"parameter\":\"intercept a\",\"q025\":0.950807,\"q25\":0.995074,\"q75\":1.04091,\"q975\":1.08237,\"sd\":0.0336806,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":1.04474,\"median\":1.04495,\"method\":\"streaming SVI\",\"parameter\":\"intercept a\",\"q025\":1.0159,\"q25\":1.03502,\"q75\":1.05489,\"q975\":1.07243,\"sd\":0.0145584,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.412736,\"median\":0.412611,\"method\":\"NUTS\",\"parameter\":\"slope b\",\"q025\":0.361623,\"q25\":0.395364,\"q75\":0.430053,\"q975\":0.463544,\"sd\":0.0259576,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.386244,\"median\":0.385755,\"method\":\"streaming SVI\",\"parameter\":\"slope b\",\"q025\":0.360953,\"q25\":0.377767,\"q75\":0.395082,\"q975\":0.410425,\"sd\":0.0127558,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.498863,\"median\":0.498127,\"method\":\"NUTS\",\"parameter\":\"sd of site intercepts\",\"q025\":0.445241,\"q25\":0.479126,\"q75\":0.517685,\"q975\":0.556574,\"sd\":0.0283744,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.476343,\"median\":0.476075,\"method\":\"streaming SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.453334,\"q25\":0.467642,\"q75\":0.4849,\"q975\":0.500062,\"sd\":0.0124695,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.2593,\"median\":0.258958,\"method\":\"NUTS\",\"parameter\":\"sd of site slopes\",\"q025\":0.211142,\"q25\":0.241095,\"q75\":0.276084,\"q975\":0.312213,\"sd\":0.0257042,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.227336,\"median\":0.22672,\"method\":\"streaming SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.208142,\"q25\":0.220074,\"q75\":0.234137,\"q975\":0.248107,\"sd\":0.0103073,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.46715,\"median\":0.468525,\"method\":\"NUTS\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.245129,\"q25\":0.392255,\"q75\":0.544553,\"q975\":0.67821,\"sd\":0.111872,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":0.695485,\"median\":0.697221,\"method\":\"streaming SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.615353,\"q25\":0.669288,\"q75\":0.723223,\"q975\":0.767945,\"sd\":0.0398214,\"window\":\"window 2\"},{\"__src\":\"_s1\",\"mean\":1.02803,\"median\":1.02832,\"method\":\"NUTS\",\"parameter\":\"intercept a\",\"q025\":0.969063,\"q25\":1.00765,\"q75\":1.04826,\"q975\":1.08684,\"sd\":0.0299943,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":1.04357,\"median\":1.04397,\"method\":\"streaming SVI\",\"parameter\":\"intercept a\",\"q025\":1.02074,\"q25\":1.03586,\"q75\":1.0512,\"q975\":1.06395,\"sd\":0.0111035,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.411386,\"median\":0.411221,\"method\":\"NUTS\",\"parameter\":\"slope b\",\"q025\":0.370309,\"q25\":0.39672,\"q75\":0.425962,\"q975\":0.45392,\"sd\":0.0215077,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.384313,\"median\":0.38422,\"method\":\"streaming SVI\",\"parameter\":\"slope b\",\"q025\":0.365681,\"q25\":0.377584,\"q75\":0.391391,\"q975\":0.403619,\"sd\":0.00996614,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.498148,\"median\":0.497487,\"method\":\"NUTS\",\"parameter\":\"sd of site intercepts\",\"q025\":0.452551,\"q25\":0.481254,\"q75\":0.513934,\"q975\":0.548691,\"sd\":0.0244202,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.473783,\"median\":0.473554,\"method\":\"streaming SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.45632,\"q25\":0.467374,\"q75\":0.480239,\"q975\":0.492396,\"sd\":0.00951671,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.256107,\"median\":0.255871,\"method\":\"NUTS\",\"parameter\":\"sd of site slopes\",\"q025\":0.219274,\"q25\":0.242881,\"q75\":0.268896,\"q975\":0.29525,\"sd\":0.0194468,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.230392,\"median\":0.230167,\"method\":\"streaming SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.214314,\"q25\":0.225181,\"q75\":0.235628,\"q975\":0.247279,\"sd\":0.00829777,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.388203,\"median\":0.390559,\"method\":\"NUTS\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.210076,\"q25\":0.330574,\"q75\":0.448895,\"q975\":0.555103,\"sd\":0.0884139,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":0.650748,\"median\":0.65163,\"method\":\"streaming SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.57643,\"q25\":0.628889,\"q75\":0.674723,\"q975\":0.715335,\"sd\":0.034989,\"window\":\"window 3\"},{\"__src\":\"_s1\",\"mean\":1.02818,\"median\":1.02848,\"method\":\"NUTS\",\"parameter\":\"intercept a\",\"q025\":0.972863,\"q25\":1.00922,\"q75\":1.04721,\"q975\":1.08227,\"sd\":0.0281021,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":1.04619,\"median\":1.04573,\"method\":\"streaming SVI\",\"parameter\":\"intercept a\",\"q025\":1.02983,\"q25\":1.03994,\"q75\":1.05237,\"q975\":1.06505,\"sd\":0.00911938,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.427022,\"median\":0.427194,\"method\":\"NUTS\",\"parameter\":\"slope b\",\"q025\":0.388786,\"q25\":0.41402,\"q75\":0.440243,\"q975\":0.464749,\"sd\":0.019478,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.388633,\"median\":0.388603,\"method\":\"streaming SVI\",\"parameter\":\"slope b\",\"q025\":0.372419,\"q25\":0.383086,\"q75\":0.394557,\"q975\":0.404311,\"sd\":0.0082166,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.500572,\"median\":0.499945,\"method\":\"NUTS\",\"parameter\":\"sd of site intercepts\",\"q025\":0.458055,\"q25\":0.484872,\"q75\":0.515542,\"q975\":0.546748,\"sd\":0.0226061,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.475654,\"median\":0.47552,\"method\":\"streaming SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.459445,\"q25\":0.470331,\"q75\":0.480973,\"q975\":0.492239,\"sd\":0.00822664,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.266296,\"median\":0.265623,\"method\":\"NUTS\",\"parameter\":\"sd of site slopes\",\"q025\":0.233774,\"q25\":0.254896,\"q75\":0.277499,\"q975\":0.30056,\"sd\":0.0169946,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.232965,\"median\":0.23257,\"method\":\"streaming SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.220454,\"q25\":0.228442,\"q75\":0.237565,\"q975\":0.246957,\"sd\":0.00677572,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.330878,\"median\":0.3317,\"method\":\"NUTS\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.180589,\"q25\":0.28147,\"q75\":0.383316,\"q975\":0.47282,\"sd\":0.0754375,\"window\":\"window 4\"},{\"__src\":\"_s1\",\"mean\":0.606244,\"median\":0.606966,\"method\":\"streaming SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.543321,\"q25\":0.585253,\"q75\":0.627581,\"q975\":0.664582,\"sd\":0.0310151,\"window\":\"window 4\"},{\"__src\":\"_s2\",\"mean\":1.06053,\"median\":1.06011,\"method\":\"batch SVI\",\"parameter\":\"intercept a\",\"q025\":1.01848,\"q25\":1.04611,\"q75\":1.07526,\"q975\":1.10342,\"sd\":0.0220867,\"window\":\"window 1\"},{\"__src\":\"_s2\",\"mean\":0.360277,\"median\":0.360857,\"method\":\"batch SVI\",\"parameter\":\"slope b\",\"q025\":0.322299,\"q25\":0.347254,\"q75\":0.373537,\"q975\":0.397547,\"sd\":0.0194327,\"window\":\"window 1\"},{\"__src\":\"_s2\",\"mean\":0.464952,\"median\":0.464335,\"method\":\"batch SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.431585,\"q25\":0.452863,\"q75\":0.476193,\"q975\":0.501077,\"sd\":0.0179081,\"window\":\"window 1\"},{\"__src\":\"_s2\",\"mean\":0.217263,\"median\":0.216837,\"method\":\"batch SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.187177,\"q25\":0.205831,\"q75\":0.228028,\"q975\":0.251059,\"sd\":0.0163094,\"window\":\"window 1\"},{\"__src\":\"_s2\",\"mean\":0.692524,\"median\":0.698099,\"method\":\"batch SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.554934,\"q25\":0.65327,\"q75\":0.737689,\"q975\":0.800958,\"sd\":0.0635653,\"window\":\"window 1\"},{\"__src\":\"_s2\",\"mean\":1.01894,\"median\":1.01892,\"method\":\"batch SVI\",\"parameter\":\"intercept a\",\"q025\":0.991447,\"q25\":1.00932,\"q75\":1.02863,\"q975\":1.04744,\"sd\":0.0141801,\"window\":\"window 2\"},{\"__src\":\"_s2\",\"mean\":0.410774,\"median\":0.410956,\"method\":\"batch SVI\",\"parameter\":\"slope b\",\"q025\":0.386697,\"q25\":0.402594,\"q75\":0.419088,\"q975\":0.434504,\"sd\":0.0123273,\"window\":\"window 2\"},{\"__src\":\"_s2\",\"mean\":0.491952,\"median\":0.492022,\"method\":\"batch SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.466844,\"q25\":0.483316,\"q75\":0.50048,\"q975\":0.517372,\"sd\":0.0126281,\"window\":\"window 2\"},{\"__src\":\"_s2\",\"mean\":0.251338,\"median\":0.251214,\"method\":\"batch SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.232362,\"q25\":0.244319,\"q75\":0.258251,\"q975\":0.270927,\"sd\":0.0100692,\"window\":\"window 2\"},{\"__src\":\"_s2\",\"mean\":0.471192,\"median\":0.471719,\"method\":\"batch SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.387876,\"q25\":0.442758,\"q75\":0.501063,\"q975\":0.548614,\"sd\":0.042453,\"window\":\"window 2\"},{\"__src\":\"_s2\",\"mean\":1.02386,\"median\":1.02329,\"method\":\"batch SVI\",\"parameter\":\"intercept a\",\"q025\":1.00172,\"q25\":1.01611,\"q75\":1.03162,\"q975\":1.04657,\"sd\":0.0115764,\"window\":\"window 3\"},{\"__src\":\"_s2\",\"mean\":0.409331,\"median\":0.409188,\"method\":\"batch SVI\",\"parameter\":\"slope b\",\"q025\":0.389464,\"q25\":0.402339,\"q75\":0.415983,\"q975\":0.430301,\"sd\":0.0103609,\"window\":\"window 3\"},{\"__src\":\"_s2\",\"mean\":0.490604,\"median\":0.490545,\"method\":\"batch SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.469722,\"q25\":0.483369,\"q75\":0.497858,\"q975\":0.511905,\"sd\":0.0106203,\"window\":\"window 3\"},{\"__src\":\"_s2\",\"mean\":0.251535,\"median\":0.251361,\"method\":\"batch SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.236447,\"q25\":0.246067,\"q75\":0.25677,\"q975\":0.267463,\"sd\":0.00791211,\"window\":\"window 3\"},{\"__src\":\"_s2\",\"mean\":0.391655,\"median\":0.393783,\"method\":\"batch SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.321647,\"q25\":0.367841,\"q75\":0.416259,\"q975\":0.456346,\"sd\":0.03544,\"window\":\"window 3\"},{\"__src\":\"_s2\",\"mean\":1.02543,\"median\":1.02545,\"method\":\"batch SVI\",\"parameter\":\"intercept a\",\"q025\":1.00691,\"q25\":1.01906,\"q75\":1.03184,\"q975\":1.04393,\"sd\":0.00945642,\"window\":\"window 4\"},{\"__src\":\"_s2\",\"mean\":0.423535,\"median\":0.423753,\"method\":\"batch SVI\",\"parameter\":\"slope b\",\"q025\":0.407848,\"q25\":0.418242,\"q75\":0.428949,\"q975\":0.440105,\"sd\":0.00816419,\"window\":\"window 4\"},{\"__src\":\"_s2\",\"mean\":0.494571,\"median\":0.494434,\"method\":\"batch SVI\",\"parameter\":\"sd of site intercepts\",\"q025\":0.477491,\"q25\":0.488555,\"q75\":0.500468,\"q975\":0.51146,\"sd\":0.0086765,\"window\":\"window 4\"},{\"__src\":\"_s2\",\"mean\":0.262004,\"median\":0.261717,\"method\":\"batch SVI\",\"parameter\":\"sd of site slopes\",\"q025\":0.247928,\"q25\":0.257065,\"q75\":0.26689,\"q975\":0.2768,\"sd\":0.00731778,\"window\":\"window 4\"},{\"__src\":\"_s2\",\"mean\":0.326552,\"median\":0.326825,\"method\":\"batch SVI\",\"parameter\":\"correlation of intercepts and slopes\",\"q025\":0.270491,\"q25\":0.30691,\"q75\":0.34605,\"q975\":0.381917,\"sd\":0.0288892,\"window\":\"window 4\"}]},\"facet\":{\"field\":\"parameter\",\"sort\":[\"intercept a\",\"slope b\",\"sd of site intercepts\",\"sd of site slopes\",\"correlation of intercepts and slopes\"],\"type\":\"nominal\"},\"resolve\":{\"scale\":{\"y\":\"independent\"}},\"spec\":{\"height\":180,\"layer\":[{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[\"legend_selection\"],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},\"encoding\":{\"color\":{\"field\":\"method\",\"scale\":{\"domain\":[\"NUTS\",\"streaming SVI\"],\"range\":[\"#2a78d6\",\"#eb6834\"]},\"type\":\"nominal\"},\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"window\",\"type\":\"nominal\"},{\"field\":\"method\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"axis\":{\"title\":null},\"field\":\"window\",\"title\":\"data seen up to\",\"type\":\"nominal\"},\"xOffset\":{\"field\":\"method\",\"type\":\"nominal\"},\"y\":{\"field\":\"q025\",\"scale\":{\"zero\":false},\"title\":\"value (dimensionless)\",\"type\":\"quantitative\"},\"y2\":{\"field\":\"q975\"}},\"mark\":{\"strokeWidth\":1.5,\"type\":\"rule\"},\"params\":[{\"bind\":\"scales\",\"name\":\"grid\",\"select\":{\"encodings\":[\"y\"],\"type\":\"interval\"}},{\"bind\":\"legend\",\"name\":\"legend_selection\",\"select\":{\"fields\":[\"method\"],\"type\":\"point\"}}],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},\"encoding\":{\"color\":{\"field\":\"method\",\"scale\":{\"domain\":[\"NUTS\",\"streaming SVI\"],\"range\":[\"#2a78d6\",\"#eb6834\"]},\"type\":\"nominal\"},\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"window\",\"type\":\"nominal\"},{\"field\":\"method\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"axis\":{\"title\":null},\"field\":\"window\",\"title\":\"data seen up to\",\"type\":\"nominal\"},\"xOffset\":{\"field\":\"method\",\"type\":\"nominal\"},\"y\":{\"field\":\"q25\",\"scale\":{\"zero\":false},\"title\":\"value (dimensionless)\",\"type\":\"quantitative\"},\"y2\":{\"field\":\"q75\"}},\"mark\":{\"strokeWidth\":8.0,\"type\":\"rule\"},\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"window\",\"type\":\"nominal\"},{\"field\":\"method\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"field\":\"window\",\"type\":\"nominal\"},\"xOffset\":{\"field\":\"method\",\"type\":\"nominal\"},\"y\":{\"field\":\"median\",\"scale\":{\"zero\":false},\"type\":\"quantitative\"}},\"mark\":{\"color\":\"white\",\"filled\":true,\"size\":80,\"stroke\":\"#333\",\"strokeWidth\":1.5,\"type\":\"point\"},\"transform\":[{\"filter\":\"datum.__src === '_s1'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"window\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"axis\":{\"title\":null},\"field\":\"window\",\"title\":\"data seen up to\",\"type\":\"nominal\"},\"y\":{\"field\":\"q025\",\"scale\":{\"zero\":false},\"title\":\"value (dimensionless)\",\"type\":\"quantitative\"},\"y2\":{\"field\":\"q975\"}},\"mark\":{\"strokeWidth\":1.5,\"style\":\"aov-ink\",\"type\":\"rule\"},\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"window\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"axis\":{\"title\":null},\"field\":\"window\",\"title\":\"data seen up to\",\"type\":\"nominal\"},\"y\":{\"field\":\"q25\",\"scale\":{\"zero\":false},\"title\":\"value (dimensionless)\",\"type\":\"quantitative\"},\"y2\":{\"field\":\"q75\"}},\"mark\":{\"strokeWidth\":8.0,\"style\":\"aov-ink\",\"type\":\"rule\"},\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]},{\"_aovLegend\":{\"dimmed\":true,\"filtered\":false,\"opacity\":null,\"params\":[],\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]},\"encoding\":{\"opacity\":{\"condition\":{\"test\":{\"and\":[{\"or\":[\"!isValid(datum[\\\"method\\\"])\",{\"empty\":true,\"param\":\"legend_selection\"}]}]},\"value\":1},\"value\":0.15},\"tooltip\":[{\"field\":\"median\",\"title\":\"value (dimensionless) (estimate)\",\"type\":\"quantitative\"},{\"field\":\"window\",\"type\":\"nominal\"},{\"field\":\"q025\",\"title\":\"q025\",\"type\":\"quantitative\"},{\"field\":\"q975\",\"title\":\"q975\",\"type\":\"quantitative\"}],\"x\":{\"field\":\"window\",\"type\":\"nominal\"},\"y\":{\"field\":\"median\",\"scale\":{\"zero\":false},\"type\":\"quantitative\"}},\"mark\":{\"color\":\"white\",\"filled\":true,\"size\":80,\"stroke\":\"#333\",\"strokeWidth\":1.5,\"type\":\"point\"},\"transform\":[{\"filter\":\"datum.__src === '_s2'\"}]}],\"width\":220},\"title\":\"Population estimates as sites and visits arrive\",\"description\":\"Population medians with 50 % and 95 % intervals: blue NUTS, orange streaming SVI, neutral ink batch SVI. All methods use exactly the observations seen so far.\"}","caption":"Population medians with 50 % and 95 % intervals: blue NUTS, orange streaming SVI, neutral ink batch SVI. All methods use exactly the observations seen so far.","alt":"Population medians with 50 % and 95 % intervals: blue NUTS, orange streaming SVI, neutral ink batch SVI. All methods use exactly the observations seen so far."}| window | sites | new counts | counts so far | streaming fit | iterations | checkpoint | batch SVI on everything | NUTS on everything |
|---|---|---|---|---|---|---|---|---|
| 1 | 300 | 600 | 600 | 42 s | 7400 | 83.3 kB | 6 s | 61 s |
| 2 | 340 | 680 | 1280 | 7 s | 7100 | 92.3 kB | 9 s | 65 s |
| 3 | 380 | 760 | 2040 | 7 s | 7100 | 101 kB | 7 s | 61 s |
| 4 | 400 | 800 | 2840 | 6 s | 7100 | 106 kB | 7 s | 1.9 min |
Each update minibatches only the units of its window; the checkpoint grows with the number of distinct sites. The narrow global sds carry over, and the sequential approximation also drifts away from the accumulated-data posterior.
At the final window the correlation mean is 3.7 NUTS posterior sds from its NUTS estimate. Across the window references, maximum R-hat is 1.005, minimum bulk/tail ESS are 888/1083, and 0 retained transitions diverged.
Start from fits on slices of the data
brm_svi_combine turns checkpoints from disjoint slices of units into a start for the full fit. The global factors multiply as Gaussian sites with the prior counted once; each unit takes its slice's factor. The full model remains the target. Name parameters in exclude when their meaning changes between a slice and the full model; those parameters start cold.
# Fits of four slices of the sites, their checkpoints combined into a start for the
# full fit. `full` is prepared once, so the combination and the fit share it.
function combined_start_fit(data, slices; progress=nothing)
slice_fits = [brm_svi_fit(site_counts_model(slice); batch_size=32, rng=Xoshiro(s), progress)
for (s, slice) in enumerate(slices)]
full = brm_nlme_model(RKBRMI(site_counts_model(data)); ad_backend=AutoEnzyme(; mode=Enzyme.Reverse))
start = brm_svi_combine(full, brm_svi_checkpoint.(slice_fits))
started = brm_svi_fit(full; start, batch_size=32, rng=Xoshiro(1), ndraws=2000, progress)
(; full, start, started)
end| start of the full fit | optimization | total including slices | iterations | of them exploring | site gradients | largest population mean error (NUTS sds) |
|---|---|---|---|---|---|---|
| cold start | 44 s | 3.2 min | 7700 | 4600 | 985600 | 0.36 |
| start from four slices | 11 s | 74 s | 7100 | 4000 | 908800 | 0.28 |
The combined start shortens exploration here. Total times include preparation and compilation, depend on run order, and are not a speed benchmark.
Two-stage fits and Laplace
brm_two_stage_fit, through the NLMEEstimation extension, estimates the population by SAEM and samples each unit's conditional posterior at those fixed estimates. It does not propagate population uncertainty.
brm_svi_fit with laplace=GaussianBlockLaplace() integrates a shared Gaussian block jointly with Gaussian population coefficients and unit effects, conditional on the remaining variational globals. It requires affine Gaussian likelihoods and visits the whole window at each update.
Reproduce
The research recipe contains the synthetic data generator, fit commands, cache rules and figure source. Fits run once locally; the documentation build renders their summary tables and AlgebraOfVega specifications. It never samples.
| receipt | value |
|---|---|
| recipe | site-counts-v2-quorum |
| commit | dab3a94d0e0d0bae4d2a79d9486880072a593851 |
| uncommitted_changes | true |
| started | 2026-11-11T00:31:41.967 |
| host | a Linux fleet host |
| Julia | 1.10.12 |
| BayesianRegressionModels | 0.2.272 |
| WarmupHMC | 0.2.11 |
| Enzyme | 0.13.218 |
| threads | 1 |
The complete receipt also records settings, seeds, input keys and fit source hashes.