Covariate-moderated Empirical Bayes Matrix Factorization Methods

[Up] [Top]

Documentation for package ‘comoR’ version 0.0.1.03

Help Pages

.converged check if ELBO is converged to some tolerance threshold
.diff helper function gets difference in ELBO between iterations
autoselect_scales_mix_exp adapted from autoselect.mxisqp try to select a default range for the sigmaa values that should be used, based on the values of betahat and sebetahat mode is the location about which inference is going to be centered gridmult is the multiplier by which the sds differ across the grid
autoselect_scales_mix_norm adapted from autoselect.mxisqp try to select a default range for the sigmaa values that should be used, based on the values of betahat and sebetahat mode is the location about which inference is going to be centered gridmult is the multiplier by which the sds differ across the grid
cal_ind_moment12_mix_norm Compute individual posterior first and second moment
cal_purity Cal purity ouput como2
cEBMF covariate moderated Empirical Bayes Matrix Factorization
como_prep_data Prepare data for multinomial regression
compute_data_loglikelihood Compute data likelihood
compute_elbo Compute ELBO, but if there is not ELBO
compute_elbo.default Default no ELBO
compute_posterior_assignment Compute Posterior Assignment Probabilities For each data point return posterior assignment probabilities
convolved_logpdf Convolved logpdf
convolved_logpdf.exp Convolution between an exponential and Normal distribution with variance given by sum of component dist. and noise
convolved_logpdf.normal Normal distribution with variance given by sum of component dist. and noise
convolved_logpdf.point Just a normal distribution with mean centered at the point mass
data_initialize_como data_initialize_como @param data description @param scales description @param mu0 description @param var0 description @param nullweight description @param mnreg_type description @param param_nnet description
get_fdr Compute individual fdr value como/como2 model with centered normal mixture
get_fdr.como Compute individual fdr value como model with centered normal mixture
get_lfdr Compute individual lfdr value como/como2 model with centered normal mixture
initialize_como Function implementation the como mode
init_cEBMF Initialize cEBMF object
out_prep.cEBMF prepare output of cEMBF function
post_mean_sd Compute post first and second moment from como and como2 object
post_mean_sd.como Compute individual posterior mean and sd under como model
set_xi set xi, update tau
sigmoid Sigmoid sigmoid function coverts log-odds to probability
sim_como_beta Simulate data under como model with beta distribution
sim_mixture_beta Simulate mixture of beta distribution
sim_ser_with_covariates simulate SER with 3 covariates
sim_susie Simulate data under the logistic SuSiE model
sim_twococomo Simulate data under the twococomo SuSiE model
t_ind_var Compute individual posterior variance from marginal normal mean model
t_ind_var.como Compute individual posterior variance from marginal normal mean model
update_params Update Params
update_params.normal Note: only updates the variance parameter, these components are assumed to be mean 0
update_tau.cEBMF Estimate noise value