Hello,
In a dataset is it possible to choose for some rows to be in the train or the test.
Example :
Data 10 rows
I want row 9 to be in the test not in the train.
Thanks
Hello @Brain,
you can not do such “row labeling” in Partitioning node but you can use Rule-based Row Filter/Splitter to prepare train and test dataset to your needs/wishes.
Br,
Ivan
2 Likes
Thanks
Please can you show me how to configure the node.
10 rows
Partinioning 70/30
i want to have row 3 and row 4 in the train set.
Thanks for your help
MISSING VALUE.xlsx (10.5 KB)
Not sure what your data looks like and all other requirements but
You could do a rule engine rowindex 3 or 4 → 10000
Then sort your data on that column and take the first 70 percent as training data (rowsampling or partitioning node and take from top)
br
3 Likes
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.