Need to do stratified sampling in a recipe Sample/Filter

[Yasmine]

Hello,

I created a Recipe Sample/Filter on Dataiku and I need to do a stratified sampling on my dataset. I found that this type of sampling is only available when you explore the data.

Is this type of sampling available on Dataiku? If not, how to do it in Python code please?

Thank you in advance for your return,
Kind regards

[pmasiphelps · Accepted solution]

Hi,

Check out this other post: https://community.dataiku.com/t5/Using-Dataiku/Split-dataset-by-stratified-sampling/m-p/2151

It has python examples in the comments 🙂

Best,

Pat

[Marlan]

Hi @Yasmine,

FYI, I submitted a product idea to provide an easy stratify option for train/test splits in the visual ML tool:

https://community.dataiku.com/t5/Product-Ideas/Easy-quot-stratify-quot-option-at-train-test-split-st...

Consider voting for it if it would be useful to you.

Regards,

Marlan