scaleFreeFitIndex: Calculation of fitting statistics for evaluating scale free...

In nosarcasm/WGCNA: Weighted Correlation Network Analysis

Description

The function scaleFreeFitIndex calculates several indices (fitting statistics) for evaluating scale free topology fit. The input is a vector (of connectivities) k. Next k is discretized into nBreaks number of equal-width bins. Let's denote the resulting vector dk. The relative frequency for each bin is denoted p.dk.

Usage

scaleFreeFitIndex(k, nBreaks = 10, removeFirst = FALSE)

Arguments

k	numeric vector whose components contain non-negative values
nBreaks	positive integer. This determines the number of equal width bins.
removeFirst	logical. If TRUE then the first bin will be removed.

Value

Data frame with columns

Rsquared.SFT	the model fitting index (R.squared) from the following model lm(log.p.dk ~ log.dk)
slope.SFT	the slope estimate from model lm(log(p(k))~log(k))
truncatedExponentialAdjRsquared	the adjusted R.squared measure from the truncated exponential model given by lm2 = lm(log.p.dk ~ log.dk + dk).

Author(s)

Steve Horvath

nosarcasm/WGCNA documentation built on May 28, 2019, 1:01 p.m.