Add Guassian Mixture Models to available clustering models

[nathaniel-johnson22]

install.packages(c("reprex", "sessioninfo"), repos = "http://cran.r-project.org")

Gaussian Mixture

I would like the ability to run gaussian mixture models using tidyclust. GMM allows clusters to be extracted assuming data is drawn from multiple gaussian distributions. Common implementations in R include the mclust

Specification Example

gmm_spec <- gaussian_mixture() %>% parsnip::set_engine("Mclust")

Fit and Prediction would work similar to kmeans, mclust produces both discrete classification (model$classification) and probabilities (a matrix: model$z)

Fit and Predict Example

gmm_fit <- gmm_spec %>% fit(~ bill_length_mm + bill_depth_mm, data = penguins )

new_penguin <- tibble(bill_length_mm = 42,  bill_depth_mm = 17)

kmeans_fit %>%  predict(new_penguin)

Currently, parsnip does not allow custom-built unsupervised models. An alternative would be a similar framework in tidyclust that extends parsnip to allow this.

[nathaniel-johnson22]

This is a duplicate of #5

[github-actions]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.