permutationMC {OmicsLonDA} | R Documentation |
Permutes the group label of the samples in order to construct the testStatistics empirical distibution
permutationMC(formula = Count ~ Time, perm.dat = NULL, n.perm = 500, fit.method = "ssgaussian", points, parall = FALSE, prefix = "Test")
formula |
formula to be passed to the regression model |
perm.dat |
dataframe has the Count, Group, Subject, Time |
n.perm |
number of permutations |
fit.method |
The fitting method (ssgaussian) |
points |
The points at which the prediction should happen |
parall |
boolean to indicate whether to use multicore. |
prefix |
prefix to be used to create directory for the analysis results |
a list of the fitted model for each group for all the permutations
Ahmed Metwally (ametwall@stanford.edu)
library(SummarizedExperiment) data("omicslonda_data_example") omicslonda_se_object_adjusted = adjustBaseline( se_object = omicslonda_data_example$omicslonda_se_object) omicslonda_test_object = omicslonda_se_object_adjusted[1,] se_object = omicslonda_test_object dt = data.frame(colData(se_object)) dt$Count = as.vector(assay(se_object)) Group = as.character(dt$Group) group.levels = sort(unique(Group)) gr.1 = as.character(group.levels[1]) gr.2 = as.character(group.levels[2]) df = dt levels(df$Group) = c(levels(df$Group), "0", "1") df$Group[which(df$Group == gr.1)] = 0 df$Group[which(df$Group == gr.2)] = 1 group.0 = df[df$Group == 0, ] group.1 = df[df$Group == 1, ] points = seq(100, 130) perm = permutationMC(formula = Count ~ Time, perm.dat = df, n.perm = 10, fit.method = "ssgaussian", points = points, parall = FALSE, prefix = tempfile())