i tried tprincipal example - the result doesnot precisely correspond to the declared one - root.cern.ch/root/html/examples/principal.C.html - it might be due to non-zeroing of data array. BUT, it seems to me, that MakeHistograms doesnot work for all but ‘e’ option, MakeCode() rewrites the original macro file and - this generated code doesnot work properly - at least X2P function (at least i would expect some other results…).
Well it is 2000 code.
Please be more precise :
- which root version, which platform ?
- what do you mean by the result does not precisely corespond
to the declared one ?
- generated code does not work properly ?
- 2000 code ?
two apologies -
for the first - I started with the macro found on the inet, there is data[n - c + j]=0 already in tutorial.
the second - default MakeCode() filename was ‘pca.C’ which rewrote my code, the tutorial is named different. It was not a problem at all.
There are histogram making options SDX, that doesnot seem to work (’ E’ is the only one working) ( principal->MakeHistograms(“pcas”,“s”)
And I dont exactly understand what X2P does. It seems that it transforms differences from mean values to something (differences from mean values
of dependent parameters?, but the results doesnot look like.). Do you have some comment on it please?
root 4.0.4/00 1May2005, Fedora 3
cint 5.15.169, Mar 14 2005