High-Dimensional Spatial Covariate-Augmented Overdispersed Poisson Factor Model
================================================================
We introduce an efficient latent representation learning approach tailored specifically for high-dimensional, large-scale spatial count data, incorporating additional covariates for enhanced performance. To model correlations among variables measured at a shared spatial location, we utilize a covariate-augmented overdispersed Poisson factor model. We distinguish between high-dimensional covariates sharing similar attributes and those serving as control variables to enrich the representation learning process. To capture the spatial dependency of each variable across different locations, we apply a conditional autoregressive model to the latent factors. Furthermore, we propose a variational expectation-maximization algorithm to estimate the model parameters and latent factors, imposing a low-rank constraint on the high-dimensional regression coefficient matrix.
Check out Package Website for a more complete description of the methods and analyses.
“SpaCOAP” depends on the ‘Rcpp’ and ‘RcppArmadillo’ package, which requires appropriate setup of computer. For the users that have set up system properly for compiling C++ files, the following installation command will work.
## Method 1:
if (!require("remotes", quietly = TRUE))
install.packages("remotes")
remotes::install_github("feiyoung/SpaCOAP")
## Method 2: install from CRAN
install.packages("SpaCOAP")
For usage examples and guided walkthroughs, check the vignettes directory of the repo.
For the codes in simulation study, check the simu_code directory of the repo.
SpaCOAP version 1.2 released! (2024-05-25)