It's my first time doing classification, sorry in advance for my ignorance.
I have some data and it seems easy to split into train and test sets with Pandas.
Is there a way to split into train, test and evaluation sets instead?
Asked
Active
Viewed 567 times
0
-
You can do this in `sklearn` using `sklearn.cross_validation.train_test_split`. The duplicate post that this was linked to was for using numpy/pandas. http://scikit-learn.org/stable/modules/generated/sklearn.cross_validation.train_test_split.html – ode2k Sep 15 '16 at 00:20
-
@ode2k thanks for your comment, isn't that the very same function my linked question uses? – gsamaras Sep 15 '16 at 00:21
-
Actually, the answer in the linked question uses numpy to split into the 3 sets that you are asking for. I guess sklearn only does the training / test sets, not evaluation set. So if you are able to use numpy for splitting them up, that is a good solution. – ode2k Sep 15 '16 at 00:29
-
Yeah I think I will go for it, thank you @ode2k! – gsamaras Sep 15 '16 at 00:46