Easy Visualization of Conditional Effects from Regression Models

Offers a flexible and user-friendly interface for visualizing conditional effects from a broad range of regression models, including mixed-effects and generalized additive (mixed) models. Compatible model types include lm(), rlm(), glm(), glm.nb(), betareg(), and gam() (from 'mgcv'); nonlinear models via nls(); generalized least squares via gls(); and survival models via coxph() (from 'survival'). Mixed-effects models with random intercepts and/or slopes can be fitted using lmer(), glmer(), glmer.nb(), glmmTMB(), or gam() (from 'mgcv', via smooth terms). Plots are rendered using base R graphics with extensive customization options. Approximate confidence intervals for nls() and betareg() models are computed using the delta method. Robust standard errors for rlm() are computed using the sandwich estimator (Zeileis 2004) . For beta regression using 'betareg', see Cribari-Neto and Zeileis (2010) . For mixed-effects models with 'lme4', see Bates et al. (2015) . For models using 'glmmTMB', see Brooks et al. (2017) . Methods for generalized additive models using 'mgcv' follow Wood (2017) .

Automate the Creation of Generalized Additive Models (GAMs)

This wrapper package for 'mgcv' makes it easier to create high-performing Generalized Additive Models (GAMs). With its central function autogam(), by entering just a dataset and the name of the outcome column as inputs, 'AutoGAM' tries to automate the procedure of configuring a highly accurate GAM which performs at reasonably high speed, even for large datasets.

Visualization of Spline Effects in GAM and GLM Models

Creates 'ggplot2'-based visualizations of smooth effects from GAM (Generalized Additive Models) fitted with 'mgcv' and spline effects from GLM (Generalized Linear Models). Supports survey-weighted models ('svyglm', 'svycoxph') from the 'survey' package, interaction terms, and provides hazard ratio plots with histograms for survival analysis. Wood (2017, ISBN:9781498728331) provides comprehensive methodology for generalized additive models.

Visualising Statistical Models using Dynamic Nomograms

Demonstrate the results of a statistical model object as a dynamic nomogram in an RStudio panel or web browser. The package provides two generics functions: DynNom, which display statistical model objects as a dynamic nomogram; DNbuilder, which builds required scripts to publish a dynamic nomogram on a web server such as the < https://www.shinyapps.io/>. Current version of 'DynNom' supports stats::lm, stats::glm, survival::coxph, rms::ols, rms::Glm, rms::lrm, rms::cph, and mgcv::gam model objects.

Estimate Diet Proportions Using Multivariate Tweedie Model

Defines predict function that transforms output from a Tweedie Generalized Linear Mixed Model (using 'glmmTMB'), Generalized Additive Model (using 'mgcv'), or spatio-temporal Generalized Linear Mixed Model (using package 'tinyVAST'), and returns predicted proportions (and standard errors) across a grouping variable from an equivalent multivariate-logit Tweedie model. These predicted proportions can then be used for standard plotting and diagnostics. See Thorson et al. 2022 .

Multivariate Spatio-Temporal Models using Structural Equations

Fits a wide variety of multivariate spatio-temporal models with simultaneous and lagged interactions among variables (including vector autoregressive spatio-temporal ('VAST') dynamics) for areal, continuous, or network spatial domains. It includes time-variable, space-variable, and space-time-variable interactions using dynamic structural equation models ('DSEM') as expressive interface, and the 'mgcv' package to specify splines via the formula interface. See Thorson et al. (2025) for more details.

Generalized Kernel Regularized Least Squares

Kernel regularized least squares, also known as kernel ridge regression, is a flexible machine learning method. This package implements this method by providing a smooth term for use with 'mgcv' and uses random sketching to facilitate scalable estimation on large datasets. It provides additional functions for calculating marginal effects after estimation and for use with ensembles ('SuperLearning'), double/debiased machine learning ('DoubleML'), and robust/clustered standard errors ('sandwich'). Chang and Goplerud (2024) provide further details.

Distributional Stochastic Frontier Analysis

Framework to fit distributional stochastic frontier models. Casts the stochastic frontier model into the flexible framework of distributional regression or otherwise known as General Additive Models of Location, Scale and Shape (GAMLSS). Allows for linear, non-linear, random and spatial effects on all the parameters of the distribution of the output, e.g. effects on the production or cost function, heterogeneity of the noise and inefficiency. Available distributions are the normal-halfnormal and normal-exponential distribution. Estimation via the fast and reliable routines of the 'mgcv' package. For more details see .

Predict and Visualize Population-Level Changes in Allele Frequencies in Response to Climate Change

Methods () are provided of calibrating and predicting shifts in allele frequencies through redundancy analysis ('vegan::rda()') and generalized additive models ('mgcv::gam()'). Visualization functions for predicted changes in allele frequencies include 'shift.dot.ggplot()', 'shift.pie.ggplot()', 'shift.moon.ggplot()', 'shift.waffle.ggplot()' and 'shift.surf.ggplot()' that are made with input data sets that are prepared by helper functions for each visualization method. Examples in the documentation show how to prepare animated climate change graphics through a time series with the 'gganimate' package. Function 'amova.rda()' shows how Analysis of Molecular Variance can be directly conducted with the results from redundancy analysis.

Meta-Analysis of Generalized Additive Models

Meta-analysis of generalized additive models and generalized additive mixed models. A typical use case is when data cannot be shared across locations, and an overall meta-analytic fit is sought. 'metagam' provides functionality for removing individual participant data from models computed using the 'mgcv' and 'gamm4' packages such that the model objects can be shared without exposing individual data. Furthermore, methods for meta-analysing these fits are provided. The implemented methods are described in Sorensen et al. (2020), , extending previous works by Schwartz and Zanobetti (2000) and Crippa et al. (2018) .