1
$\begingroup$

I was thinking of adding another column called "Test / Train" in which a 0 would indicate test and a 1 would indicate train data, and then as I'm collecting data end the train data collection phase and start entering each row with 1 in that column. Then using select rows widget to select all rows with "Test / Train" = 0 or 1. However, under PCA, the original columns go away so this wouldn't work!

How then, can I just select all rows with row id > k? I need this because the past data is supposed to predict a smaller set of future data, so I want to be sure that it works and not just select the test set randomly from the training data (which is an option in Test & Score).

$\endgroup$
1
$\begingroup$

Actually, I can add in a string column and should still be able to do the row selection trick even with PCA. In the file widget you set that column as text / meta and it shows up in the forward row selection widget.

I think this will work.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.