plotClassifierROC: Diagnostic plots for pathClassifier.

In ahmohamed/NetPathMiner: NetPathMiner for Biological Network Construction, Path Mining and Visualization

Diagnostic plots for pathClassifier.

Description

Diagnostic plots for pathClassifier.

Usage

plotClassifierROC(mix)

Arguments

mix	The result from pathClassifier.

Value

Diagnostic plots of the result from pathClassifier. itemTopROC curves for the posterior probabilities (mix\$posterior.probs) and for each HME3M component (mix\$h). This gives information about what response label each relates to. A ROC curve with an AUC < 0.5 relates to y = 0. Conversely ROC curves with AUC > 0.5 relate to y = 1. itemBottomThe likelihood convergence history for the HME3M model. If the parameters alpha or lambda are set too large then the likelihood may decrease.

Author(s)

Timothy Hancock and Ichigaku Takigawa

See Also

Other Path clustering & classification methods: pathClassifier, pathCluster, pathsToBinary, plotClusterMatrix, plotPathClassifier, plotPathCluster, predictPathClassifier, predictPathCluster

Other Plotting methods: colorVertexByAttr, layoutVertexByAttr, plotAllNetworks, plotClusterMatrix, plotCytoscapeGML, plotNetwork, plotPathClassifier, plotPaths

ahmohamed/NetPathMiner documentation built on May 15, 2023, 12:53 p.m.