Multinomial Logistic-Normal Models (really fast)
its a little tar-ball of joy
Silverman, JD, Roche, K, Holmes, ZC, David, LA, and Mukherjee, S. Journal of Machine Learning Research. 23(7), 2022:1−42.
All source code freely availale under GPL-3 License.
install.packages("fido")Or to download the development version from GitHub:
devtools::install_github("jsilve24/fido", ref="develop")A few notes:
- As of version 1.0.0, fido is now on CRAN.
- There are a few installation options that can greatly speed fido up (often by as much as 10-50 fold). For a more detailed description of installation, take a look at the installation page.
- Vignettes are prebuilt on the fido webpage. If you
want vignettes to build locally during package installation you must also pass the
build=TRUEandbuild_opts = c("--no-resave-data", "--no-manual")options toinstall_github.
- Introduction to fido::Pibble
- Non-linear models with fido::basset
- Joint Modeling (e.g., Multinomics) with fido::Orthus
- Tips for Specifying Priors
- Mitigating PCR bias
- Silverman et al., Bayesian Multinomial Logistic Normal Models through Marginally Latent Matrix-T Processes
- Silverman et al., Measuring and Mitigating PCR Bias in Microbiome Data
- Holmes et al., Short-Chain Fatty Acid Production by Gut Microbiota from Children with Obesity Differs According to Prebiotic Choice and Bacterial Community Composition
- Silverman et al., Using Influenza surveillance ntworks to estimate state-specific prevalance of SARS-CoV-2 in the United States
Have you checked out our FAQ?
I appreciate bug reports and feature requests. Please post to the github issue tracker here.