View source: R/ranger_crossRF_plot_util.R
id_non_spcf_markers | R Documentation |
Non-specific features across datasets (at least present in two of datasets): Last update: 20190130
id_non_spcf_markers( feature_res, positive_class = "disease", other_class = "health", p.adj.method = "BH", outdir = NULL )
feature_res |
the inheritant output from the function of plot.res_list, rf_clf.comps.summ or rf_clf.by_dataset.summ |
positive_class |
A string indicates one class in the 'c_category' column of metadata. |
other_class |
A string indicates the other class in the factor, such as 'health'. |
p.adj.method |
The p-value correction method, default is "bonferroni". |
outdir |
The outputh directory, default is "./". |
...
Shi Huang
ranger
df <- data.frame(rbind(t(rmultinom(15, 75, c(.21,.6,.12,.38,.099))), t(rmultinom(15, 75, c(.011,.6,.22,.28,.289))), t(rmultinom(15, 75, c(.091,.6,.32,.18,.209))), t(rmultinom(15, 75, c(.001,.6,.42,.58,.299))))) # A dataset with no feature significantly changed in all sub-datasets! df0 <- data.frame(t(rmultinom(60, 300,c(.001,.6,.2,.3,.299)))) f=factor(c(rep("A", 15), rep("B", 15), rep("C", 15), rep("D", 15))) comp_group="A" res <-rf_clf.comps.summ(df, f, comp_group, verbose=FALSE, ntree=500, p_cutoff=0.05, p.adj.method = "bonferroni", q_cutoff=0.05, outdir=NULL) feature_res<-res$feature_res id_non_spcf_markers(feature_res, positive_class="disease", other_class="health", p.adj.method= "BH", outdir=NULL)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.