gllvm

gllvm implements Generalized Linear Latent Variable Models (GLLVMs) for multivariate analysis of ecological community data, enabling dimension reduction while accounting for residual correlation between species through latent variables.

Key Features:

• Fast Approximate Estimation: Performs estimation via fast approximate likelihood-based techniques including extended variational approximation applicable to almost any combination of response type and link function.
• Model-Based Ordination: Implements constrained and concurrent ordination methods for community ecology datasets.
• Nested/Hierarchical Sampling Designs: Accounts for nested and hierarchical sampling designs in ecological data.
• Phylogenetic Random Effects: Incorporates phylogenetic random effects and other random effects structures.
• Expanded Response Types: Handles a greatly expanded range of response types for multivariate ecological data.
• Inference and Prediction: Provides enhanced capabilities for GLLVM inference, selection, and prediction.
• Joint Species Distribution Models: Serves as a general framework for joint modeling of community ecology datasets as joint species distribution models (JSDMs).

Scientific Applications:

• Community Ecology Analysis: Analyzes multivariate ecological community data across a wide range of data types.
• Dimension Reduction in Ecology: Performs dimension reduction via latent variables while accounting for residual correlation between species in ecological communities.
• Joint Species Distribution Modeling: Models joint species distributions to understand co-occurrence patterns and community assembly processes.
• Phylogenetically-Informed Analysis: Incorporates phylogenetic information into multivariate ecological analyses through phylogenetic random effects.

Methodology:

Estimation is performed via fast approximate likelihood-based techniques including extended variational approximation. Models include a small number of latent variables to perform dimension reduction while accounting for residual correlation between species.