succotashr: 'succotashr': An 'R' package for running Surrogate and...

Description succotashr functions

Description

This package contains functions for implementing the SUCCOTASH method in R. This is method to account for hiddent confounders when performing linear regression. The important functions in succotashr are succotash, succotash_given_alpha, and factor_mle.

succotashr functions

draw_beta: Draw from a mixture of normals.

factor_mle: Regularized maximum likelihood factor analysis with heteroscedastic columns.

f_val: Regularized normal log-likelihood.

lfdr_to_q: Transform local false discovery rates to q-values.

succotash: Surrogate and Confounder Correction Occuring Together with Adaptive SHrinkage.

succotash_em: An EM algorithm for maximizing the SUCCOTASH log-likelihood.

succotash_fixed: A fixed-point iteration of the EM algorithm.

succotash_given_alpha: Maximize the SUCCOTASH log-likelihood and return posterior summaries.

succotash_llike: The SUCCOTASH log-likelihood.

succotash_summaries: Provides posterior summaries in the SUCCOTASH model.

update_A: Update low rank mean in regularized maximum likelihood factor analysis.

update_sigma: Update variances in regularized maximum likelihood factor analysis.

trim: Truncates small numbers to 0.


dcgerard/succotashr documentation built on May 15, 2019, 1:25 a.m.