Package {inlaops}

Title:	Deployment, Versioning and Scoring Tools for 'inlabru' Models
Version:	0.1.2
Description:	Serves Bayesian spatial and latent Gaussian models fit with 'inlabru' as versioned Representational State Transfer (REST) application programming interfaces (APIs). Includes a 'vetiver' bridge (versioning, REST API serving with posterior uncertainty, 'Docker' packaging) for 'bru' objects, a 'parsnip' model constructor scoped to fixed-effects models, and predictive scoring helpers (continuous ranked probability, log, squared-error and Dawid-Sebastiani scores) computed from inlabru::generate() posterior samples via 'scoringRules'.
License:	GPL (≥ 3)
URL:	https://github.com/novica/inlaops, https://novica.github.io/inlaops/
BugReports:	https://github.com/novica/inlaops/issues
Additional_repositories:	https://inla.r-inla-download.org/R/stable
Encoding:	UTF-8
Imports:	dplyr, inlabru (≥ 2.15.0), rlang, tibble
Suggests:	broom, callr, fmesher, ggplot2, hardhat, httpuv, httr2, INLA, jsonlite, knitr, parsnip, pins, plumber, purrr, recipes, rsample, scoringRules, sf, sn, testthat (≥ 3.0.0), vetiver, yardstick, quarto
Enhances:	sp
VignetteBuilder:	quarto
Config/testthat/edition:	3
Config/roxygen2/version:	8.1.0
NeedsCompilation:	no
Packaged:	2026-10-21 06:00:07 UTC; novica
Author:	Novica Nakov [aut, cre, cph]
Maintainer:	Novica Nakov <nnovica@gmail.com>
Repository:	CRAN
Date/Publication:	2026-11-03 13:10:02 UTC

Cross-validation predictions for fixed-effects inlabru models

Description

Wraps rsample::vfold_cv() and re-fits the model on each training fold, returning a tidy tibble of holdout predictions. Scoped to models with fixed effects and IID random intercepts only.

Usage

bru_cv_predict(
  formula,
  family,
  data,
  v = 5L,
  pred_formula,
  n_samples = 500L,
  seed = NULL,
  strata = NULL,
  ...
)

Arguments

Details

Structured random effects (SPDE, BYM2, AR1, RW1/RW2) cannot be cross-validated by simply subsetting the data. Each requires per-fold reconstruction of the index structure: SPDE models need a new A-matrix (mesh projector) built from the fold's observation coordinates; BYM2 areal models need the adjacency graph to account for areas with no fold observations; AR1/RW models assume a contiguous integer time index that breaks when folds remove time points. None of this can be done generically. For those models use spatial block cross-validation and evaluate with bru_score().

Folds run sequentially. Callers who want parallel folds should compose with their own parallel backend (e.g. future.apply::future_lapply) around a per-fold inlabru::bru() call — this package stays out of the daemon lifecycle to keep the scope focused on bridging, not execution.

Value

A tibble with columns .truth, .pred, .pred_low, .pred_high, .pred_sd.

Examples

set.seed(1)
dat <- data.frame(x = rnorm(60))
dat$y <- 1 + 0.5 * dat$x + rnorm(60, sd = 0.5)
bru_cv_predict(
  formula = y ~ Intercept(1) + x,
  family = "gaussian",
  data = dat,
  v = 3,
  pred_formula = ~ Intercept + x,
  n_samples = 100,
  seed = 42
)

Internal fit wrapper for the parsnip inlabru engine

Description

Internal fit wrapper for the parsnip inlabru engine

Usage

bru_inlaops_fit(formula, data, dic = TRUE, waic = TRUE, options = NULL, ...)

Arguments

Value

A fitted bru object (see inlabru::bru()) with stored posterior configurations, so it can be sampled with inlabru::generate().

Examples

set.seed(1)
dat <- data.frame(x = rnorm(30))
dat$y <- 1 + 0.5 * dat$x + rnorm(30, sd = 0.5)
fit <- bru_inlaops_fit(y ~ Intercept(1) + x, data = dat, family = "gaussian")
fit$summary.fixed

Build the response-scale predict formula for the parsnip bru_reg engine

Description

The fixed-effect linear predictor wrapped in the default inverse link for the fit's family. IID random intercepts (and any other random effects) are intentionally excluded: the formula is built from summary.fixed row names, which are the fixed-effect terms only. This produces the marginal mean prediction — see the bru_reg() docs for details. Exported only so that parsnip's quoted args$formula in set_pred can reference it without :::. Not intended for end users.

Usage

bru_pred_formula(fit)

Arguments

Details

Errors if the fit specified a non-default link (via control.family = list(link = ...)) or if the family has no known default inverse link in the package's mapping. In either case, use inlabru::predict() directly.

Value

A one-sided formula.

Examples

set.seed(1)
dat <- data.frame(x = rnorm(30))
dat$y <- rpois(30, exp(0.5 + 0.3 * dat$x))
fit <- bru_inlaops_fit(y ~ Intercept(1) + x, data = dat, family = "poisson")
bru_pred_formula(fit)

Internal conf_int helper for the parsnip bru_reg engine

Description

Computes posterior credible intervals at an arbitrary level by sampling via inlabru::generate() and taking the requested quantiles. Called by parsnip's predict(type = "conf_int", level = ...) dispatch. Exported only so parsnip::set_pred's quoted func can reach it via the pkg = "inlaops" lookup; not intended for end users.

Usage

bru_predict_conf_int(object, newdata, level, formula, n.samples = 1000L)

Arguments

Value

A list with q_lower and q_upper numeric vectors.

Examples

set.seed(1)
dat <- data.frame(x = rnorm(30))
dat$y <- 1 + 0.5 * dat$x + rnorm(30, sd = 0.5)
fit <- bru_inlaops_fit(y ~ Intercept(1) + x, data = dat, family = "gaussian")
bru_predict_conf_int(
  fit,
  newdata = data.frame(x = c(-1, 0, 1)),
  level = 0.9,
  formula = bru_pred_formula(fit),
  n.samples = 200
)

parsnip model constructor for inlabru fixed-effects models

Description

Scoped to models with fixed effects and IID random intercepts only. Spatial models with BYM2, SPDE, AR1, or LGCP likelihoods should use inlabru's native interface.

Usage

bru_reg(mode = "regression", family = "gaussian")

Arguments

Details

The formula passed to parsnip::fit() must use inlabru component syntax. An explicit intercept is required: y ~ Intercept(1) + x1 + x2. Standard y ~ x1 + x2 formulas are not supported because inlabru does not add an intercept automatically.

IID random intercepts are supported in the fit, but predict() via the parsnip interface returns the marginal mean — only the fixed effects are included in the response-scale linear predictor. The IID random intercept is integrated over rather than conditioned on any specific group level, which matches the "new group" use case but silently ignores the random effect for held-out observations from a seen group. If you need conditional predictions, call inlabru::predict() directly with a formula that references the IID component.

Value

A bru_reg model spec.

Examples

bru_reg(family = "poisson")

Predictive scores for inlabru models

Description

Thin wrapper over scoringRules that takes posterior samples from inlabru::generate() and returns four predictive scores per observation.

Usage

bru_score(fit, newdata, pred_formula, y_col, n_samples = 1000L, seed = NULL)

Arguments

Details

pred_formula must produce samples of Y — including observation-model stochasticity for discrete families. The caller writes the full distributional expression. Inside an inlabru generate() formula the predictor is evaluated once per posterior sample: latent-field symbols (Intercept, SPDE field values, ...) come in as scalars, and covariates come in as vectors over observations. So ⁠length(<covariate>)⁠ gives the observation count for the per-sample ⁠r*()⁠ call:

• Poisson: ~ rpois(length(x), exp(Intercept + x))

• Binomial: ~ rbinom(length(x), size = n_trials, prob = plogis(Intercept + x))

• Gaussian: ~ rnorm(length(x), mean = Intercept + x, sd = 1 / sqrt(Precision_for_the_Gaussian_observations))

Hyperparameters are available in the formula under their standardised names (see inlabru::bru_standardise_names()), so the Gaussian observation noise is sampled jointly with the latent field.

This is the same contract inlabru::generate() already exposes — the package does not inject observation noise or dispatch on family. Users who want response-scale samples for a family not listed above write the appropriate random draw in the formula.

CRPS and the kernel log-score are strictly proper for continuous Y. For count / binomial Y the empirical CRPS on integer draws is still a reasonable proper score; scoringRules::logs_sample uses a kernel-density estimate which can be unstable on discrete supports — treat logs as approximate in that case.

Dawid-Sebastiani variance is floored at .Machine$double.eps to avoid division by zero when the sample variance collapses.

Value

A tibble with one row per observation and columns crps, logs, se, ds.

References

Gneiting, T. and Raftery, A. E. (2007). Strictly proper scoring rules, prediction, and estimation. Journal of the American Statistical Association, 102(477), 359–378. doi:10.1198/016214506000001437

Examples

set.seed(1)
dat <- data.frame(x = rnorm(40))
dat$y <- rpois(40, exp(0.5 + 0.3 * dat$x))
fit <- bru_inlaops_fit(
  y ~ Intercept(1) + x,
  data = dat[1:30, ],
  family = "poisson"
)
bru_score(
  fit,
  newdata = dat[31:40, ],
  pred_formula = ~ rpois(length(x), exp(Intercept + x)),
  y_col = "y",
  n_samples = 200,
  seed = 1
)