In hiroyukiyamamoto/loadings: Loadings for Principal Component Analysis and Partial Least Squares

knitr::opts_chunk$set(echo = TRUE)

準備

• pls-rogの関数を読み込む
• デモデータ(variable × sample)を読み込むんで、sample × variableの行列に転置する

source("https://raw.githubusercontent.com/hiroyukiyamamoto/pls-rog/master/plsrog.R")
load(file="C:/Users/yamamoto/Documents/GitHub/pls-rog/demo.Rdata")
X <- t(Z$X) # sample*variable

PLS-ROGの計算と結果

• 引数は、データ行列、群情報、平滑化パラメーターの順に設定する
• score_XがPLS-ROGのスコア
• Rはscore_Yとデータ行列Xの各変数の相関係数(PLS-ROG loading)
• Pは各列がそれぞれの成分のPLS-ROG loadingのp-value

plsrog <- plsrog(X, Z$Y, 0.999)

score_X <- plsrog[[1]] # score for X
score_Y <- plsrog[[2]] # score for Y
Wx <- plsrog[[3]] # weight vector for X
Wy <- plsrog[[4]] # weight vector for Y
R <- plsrog[[5]] # factor loading for X
P <- plsrog[[6]] # p-value of factor loading for X

plot(score_X)

hiroyukiyamamoto/loadings documentation built on May 2, 2024, 5:53 a.m.