Package: mglasso 0.1.3

mglasso: Multiscale Graphical Lasso

Inference of Multiscale graphical models with neighborhood selection approach. The method is based on solving a convex optimization problem combining a Lasso and fused-group Lasso penalties. This allows to infer simultaneously a conditional independence graph and a clustering partition. The optimization is based on the Continuation with Nesterov smoothing in a Shrinkage-Thresholding Algorithm solver (Hadj-Selem et al. 2018) <doi:10.1109/TMI.2018.2829802> implemented in python.

Authors:Edmond Sanou [aut, cre], Tung Le [ctb], Christophe Ambroise [ths], Geneviève Robin [ths]

mglasso_0.1.3.tar.gz
mglasso_0.1.3.zip(r-4.7-any)mglasso_0.1.3.zip(r-4.6-any)mglasso_0.1.3.zip(r-4.5-any)
mglasso_0.1.3.tgz(r-4.6-any)mglasso_0.1.3.tgz(r-4.5-any)
mglasso_0.1.3.tar.gz(r-4.7-any)mglasso_0.1.3.tar.gz(r-4.6-any)
mglasso_0.1.3.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
mglasso/json (API)

# Install 'mglasso' in R:

install.packages('mglasso', repos = c('https://desanou.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/desanou/mglasso/issues

Pkgdown/docs site:https://desanou.github.io

On CRAN:

4.18 score 2 stars 15 scripts 295 downloads 24 exports 76 dependencies

Last updated from:0cbbafe519. Checks:7 ERROR, 2 OK. Indexed: yes.

Target	Result	Time
linux-devel	ERROR	202
source / vignettes	OK	226
linux-release	ERROR	141
macos-release	ERROR	176
macos-oldrel	ERROR	164
windows-devel	ERROR	115
windows-release	ERROR	132
windows-oldrel	ERROR	103
wasm-release	OK	109

Exports:adj_mat conesta conesta_rwrapper configs_simu estimate_variability extract_meta fun_lines get_clusters_mgl get_mean_ROC_stat get_perf_from_raw ggplot_roc install_pylearn_parsimony mglasso mglasso_pair_param neighbor_select one_simu_extended one_simu_ROC plot_clusterpath plot_mglasso plot_res reformat_roc_res_for_ggplot seq_l2_l1_fixed sim_data symmetrize

Dependencies:askpass bit bit64 boot capushe cellranger class cli clipr corpcor cpp11 crayon curl data.table DescTools e1071 Exact expm farver forcats fs ggplot2 ggrepel gld glue gridExtra gtable haven here hms httr isoband jsonlite labeling lattice lifecycle lmom magrittr MASS Matrix mime mvtnorm openssl pillar pkgconfig png prettyunits progress proxy R.methodsS3 R.oo R.utils R6 rappdirs RColorBrewer Rcpp RcppTOML readr readxl rematch reticulate rlang rootSolve rprojroot rstudioapi S7 scales sys tibble tidyselect tzdb utf8 vctrs viridisLite vroom withr

Multiscale GLasso

Rendered frommglasso.Rmdusingknitr::knitr

Last update: 2022-09-06
Started: 2022-01-28

Citation

Development and contributors

Readme and manuals

Help Manual

Help page	Topics
Adjacency matrix	adj_mat
Init Beta 1 matrix	beta_idty
Init Beta via OLS	beta_ols
vectorize beta matrix	beta_to_vector
return precision matrix	bloc_diag
cah_glasso	cah_glasso
CONESTA solver.	conesta
CONESTA solver for numerical experiments.	conesta_rwrapper
cost function	cost
distances Beta	dist_beta
Mean error from classical regression	error
Formula from Huge paper	error_huge
TO DO: Fill upper triangular matrix then sum up with the transpose to have full matrix	expand_beta
doesn't work when dealing with matrix where diagonal of zero should be adjusted	expand_beta_deprecated
extracts meta-variables indices	extract_meta
weighted sum/difference of two regression vectors	fun_lines
Plot ROC curve and calculate AUC	get_auc
compute range of number of clusters from ROC outputs take in parameter an object from reorder_mglasso_roc_calculations	get_range_nclusters
Title	ggplot_roc
Neighborhood selection estimate	graph_estimate
Plot the image of a matrix	image_sparse
Install the python library pylearn-parsimony and other required libraries	install_pylearn_parsimony
lagrangian function	lagrangian
Check first estimate coeffs with glm	lasso_estimate
mean of randomly simulated precision matrices in the same configuration	mean_prec_mat
Merge Beta Different types of merging and their effect	merge_beta
compute clusters partition from pairs of variables to merge	merge_clusters
Merge labels	merge_labels
merge clusters from table	merge_proc
Merge X	mergeX
Inference of Multiscale Gaussian Graphical Model.	mglasso
neighbor_select	neighbor_select
One simulation configuration	one_config
compute TPR, FPR, SHD given estimated and true precision matrices	perf_one
get performances from list of estimations	perf_vec
Plot MGLasso Clusterpath	plot_clusterpath
fonction qui affiche les matrices d'adjacence à chaque niveau de la hiérarchie à automatiser utiliser niveau de legende commune	plot_mglasso
Compute precision matrix from regression vectors	precision_to_regression
pareil pour les clusters ACP dimensions ?	repart
EBIC	select_ebic_weighted
K-fold cross validation neghborhood lasso selection	select_kfold
K-fold cross validation mglasso	select_kfold_mglasso
Finds the optimal number of clusters using slope heuristic	select_partition
stability selection mglasso	select_stab_mglasso
stability selection mglasso II stars way	select_stars_mglasso
def sequences for lambda1s and lambda2s not sure if max of lambda1 s still the same as in the lasso case. But if find better equivalence will update this part	seq_l1l2
simulate data with given graph structure	sim_data
symmetrize matrix of regression vectors pxp	symmetrize