vis.res {minet}R Documentation

Visualize Results

Description

A group of functions to plot precision-recall and ROC curves and to compute f-scores from the data.frame returned by the validate function.

Usage

  pr(table)
  rates(table)
  fscores(table, beta=1)
  show.pr(table,device=-1,...)
  show.roc(table,device=-1,...)

Arguments

table This is the (steps x 5) data.frame returned by the validate function where steps is the number of thresholds used in the validation process and where columns contain TP,FP,TN,FN values (confusion matrix) as well as the threshold value used - see validate.
beta Numeric used as the weight of the recall in the f-score formula - see details. The default value of this argument is 1, meaning precision as important as recall.
device The device to be used. This parameter allows the user to plot precision-recall and receiver operating characteristic curves for various inference algorithms on the same plotting window - see examples.
... arguments passed to plot

Details

A confusion matrix contains FP,TP,FN,FP values.

  • "true positive rate" tpr = TP/(TN+TP)
  • "false positive rate" fpr = FP/(FN+FP)
  • "precision" p = TP/(FP+TP)
  • "recall" r = TP/(TP+FN)
  • "f-beta-score" Fbeta = (1+beta) * p*r/(r + beta*p)

    Value

    The function show.roc (show.pr) plots the ROC-curve (PR-curve) and returns the device associated with the plotting window.
    The function pr returns a (steps x 2) data.frame where steps is the number of thresholds used in the validation process. The first column contains precisions and the second recalls - see details.
    The function rates also returns a (steps x 2) data.frame where the first column contains true positive rates and the second column false positive rates - see details.
    The function fscores returns steps fscores according to the steps confusion matrices contained in the 'table' argument - see details.

    References

    Patrick E. Meyer, Kevin Kontos, Frederic Lafitte, and Gianluca Bontempi. Information-theoretic inference of large transcriptional regulatory networks. EURASIP Journal on Bioinformatics and Systems Biology, 2007.

    See Also

    validate, plot

    Examples

    data(syn.data)
    data(syn.net)
    # Inference
    mr <- minet( syn.data, method="mrnet", estimator="mi.empirical" )
    ar <- minet( syn.data, method="aracne", estimator="mi.empirical" )
    clr<- minet( syn.data, method="clr", estimator="mi.empirical" )
    # Validation
    mr.tbl <- validate(mr,syn.net)
    ar.tbl <- validate(ar,syn.net)
    clr.tbl<- validate(clr,syn.net)
    # Plot PR-Curves
    max(fscores(mr.tbl))
    dev <- show.pr(mr.tbl, col="green", type="b")
    dev <- show.pr(ar.tbl, device=dev, col="blue", type="b")
    show.pr(clr.tbl, device=dev, col="red",type="b")
    

    [Package minet version 1.2.0 Index]