Hm, the "sensitivity" (TP/[TP+FN]) should be equal to "recall", not the "precision". Maybe it would help if you could print the confusion matrices for a simpler binary case to track what's going on here
> On Jun 17, 2015, at 9:32 AM, Herbert Schulz <hrbrt....@gmail.com> wrote: > > Yeah i know, thats why I'm asking. i thought precision is not the same like > recall/sensitivity. > > recall == sensitivity!? > > But in this matrix, the precision is my calculated sensitivity, or is the > precision in this case the sensitivity? > > On 17 June 2015 at 15:29, Andreas Mueller <t3k...@gmail.com> wrote: > Yeah that is the rounding of using %2f in the classification report. > > > On 06/17/2015 09:20 AM, Joel Nothman wrote: >> To me, those numbers appear identical at 2 decimal places. >> >> On 17 June 2015 at 23:04, Herbert Schulz <hrbrt....@gmail.com> wrote: >> Hello everyone, >> >> i wrote a function to calculate the sensitivity,specificity, ballance >> accuracy and accuracy from a confusion matrix. >> >> >> Now i have a Problem, I'm getting different values when I'm comparing my >> Values with those from the metrics.classification_report function. >> The general problem ist, my predicted sensitivity is in the classification >> report the precision value. I'm computing every sensitivity with the one vs >> all approach. So e.g. Class 1 == true, class 2,3,4,5 are the rest (not true). >> >> I did this only to get the specificity, and to compare if i computed >> everything right. >> >> >> >> ----------- ensemble ----------- >> >> precision recall f1-score support >> >> 1.0 0.56 0.68 0.61 129 >> 2.0 0.28 0.15 0.20 78 >> 3.0 0.45 0.47 0.46 116 >> 4.0 0.29 0.05 0.09 40 >> 5.0 0.44 0.66 0.53 70 >> >> avg / total 0.43 0.47 0.43 433 >> >> >> Class: 1 >> sensitivity:0.556962025316 >> specificity: 0.850909090909 >> ballanced accuracy: 0.703935558113 >> >> Class: 2 >> sensitivity:0.279069767442 >> specificity: 0.830769230769 >> ballanced accuracy: 0.554919499106 >> >> Class: 3 >> sensitivity:0.446280991736 >> specificity: 0.801282051282 >> ballanced accuracy: 0.623781521509 >> >> Class: 4 >> sensitivity:0.285714285714 >> specificity: 0.910798122066 >> ballanced accuracy: 0.59825620389 >> >> Class: 5 >> sensitivity:0.442307692308 >> specificity: 0.927051671733 >> ballanced accuracy: 0.68467968202 >> >> >> >> >> ------------------------------------------------------------------------------ >> >> _______________________________________________ >> Scikit-learn-general mailing list >> Scikit-learn-general@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general >> >> >> >> >> ------------------------------------------------------------------------------ >> >> >> >> _______________________________________________ >> Scikit-learn-general mailing list >> >> Scikit-learn-general@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general > > > ------------------------------------------------------------------------------ > > _______________________________________________ > Scikit-learn-general mailing list > Scikit-learn-general@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general > > > ------------------------------------------------------------------------------ > _______________________________________________ > Scikit-learn-general mailing list > Scikit-learn-general@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general ------------------------------------------------------------------------------ _______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general