Possible depreciation issue with 1D gaussian mixture example

37 views
Skip to first unread message

simos kalfas

unread,
Nov 27, 2017, 9:44:28 AM11/27/17
to astroML-general

Hello!

I have been trying to fit a 1D gaussian mixture to my data and I found the example here that looked promising. Trying to run the code, I realised that the sklearn.mixture.GMM.eval() method had been depreciated (as has the GMM class but it is still usable) and replaced by the score_samples() function which has the same output. However, when I replaced it in the example code, the data structure of the input was not correct, raising an error. It seems that the new version requires a 2D array (see sklearn.utils.check_array()) which the data is run through). Is there any recommendation for fitting a 1D sample with the updated versions?

Thanks,
Simos
Reply all
Reply to author
Forward
0 new messages