MATLAB: Understanding output of perfcurve

aucdimensionperfcurveroc

[X,Y,T,AUC2,OPTROCPT] = perfcurve(classtest,classpredict,2);

My classtest as well as classpredict are 460*1, but when i run perfcurve i get x 3*1 and so it y.

I do not understand why is it so. As X,Y are used to draw ROC, I think it should give both of them as 460*1. I must be missing something.

Thank you for your time.

Best Answer

Could it be that you have relatively few distinct values of classpredict?

Related Solutions

MATLAB: Perfcurve with multiple negative classes

By default, the Y criterion is TPR computed for all possible score thresholds. The score thresholds are distinct values in the 2nd input to perfcurve - they do not depend on the negative class. TPR at a fixed threshold depends only on the scores for the positive class. That's why you get two identical columns in suby. If you want to compute two ROC curves at once, swap the X and Y criteria:

s = rand(30,1);
Y = datasample(1:3,30)';
[x,y,t,auc,optrocpt,suby,subynames] = perfcurve(Y,s,2,'negClass',[1 3],...
'ycrit','fpr','xcrit','tpr');
plot(suby(:,1),x)
hold
plot(suby(:,2),x,'r')
hold off
title('Two ROC curves')

MATLAB: Precision/Recall (perfcurve)

perfcurve() is meant for computing performance curves for your model - such as PR curves or ROC curves. In the literature, you may find these curves used for a more rigorous examination of a model's performance than can be given by a single score.

I recommend looking up the doc linked below. The Algorithms > Pointwise Confidence Bounds section explains what goes on under-the-hood after you call the function with your particular arguments. It samples from your input data multiple times, with different thresholds, to come up with a range of P/R scores, which are used to plot the curve.

perfcurve() doc: https://www.mathworks.com/help/stats/perfcurve.html

To calculate a single value, you may extract counts of TP, FP, TN & FN from your labels and predictions; then manually compute -

Precision = TP / (TP + FP)

Recall = TP / (TP + FN)

Should be done in 2 lines of code.

Alternatively, you may use confmat() to produce a confusion matrix, to readily give you TP/FP/TN/FN counts. And then do the same computation.

%Using this method, here's an anonymous function to calculate precision.
% << for an existing Confision Matrix 'ConfusionMat' >>
precision = @(confusionMat) diag(confusionMat)./sum(confusionMat,2);
% And another for recall
recall = @(confusionMat) diag(confusionMat)./sum(confusionMat,1)';
% Computes a vector of P/R values - each value corresponds to selecting a different label as +ve.

Hope it Helps!

Best Answer

Related Solutions

MATLAB: Perfcurve with multiple negative classes

MATLAB: Precision/Recall (perfcurve)

Related Question