generate_curve_data: Compute metrics needed for ROC and PR curves

In LeoStafu/metaDAF: Metagenomics tools for detecting Differentially Abundant Features

Description

Based on results from differentially abundant features (DAF) analysis and a variable used for threshold definition, it will compute 3 differents metrics, allowing to plot ROC and precision-recall curves.

Usage

generate_curve_data(dt, threshold_var = "adjusted_pvalue")

Arguments

dt	a data frame. The output of the run_DAF_analysis function. If the function return a list countaining both "raw" and "curated" results, then curated results must be used.
threshold_var	a character. The name of the column that should be used as threshold variable. By default "adjusted_pvalue".

Value

a list of 2 data frame :

roc	Countains TPR, FPR, method name and run number for each threshold value
pr	Countains TPR, precision, method name and run number for each threshold value

LeoStafu/metaDAF documentation built on July 14, 2019, 6:41 p.m.