Data generation


I have some questions about data generation with KNIME:

  1. Does any one know how to generate a data set with more categorical and numerical attributs differently correlated with each other?

  2. how can I generate a data set that have a lot of independent attributes?

  3. how to generate missing values in a lot of different columns?


Kind Regards,

Hi @Northern,
Take a look at these example workflows:

They demonstrate how you can use the modular data generation framework to generate complex data.

Hallo gab1one,

Thanks for the reply. I’ve read these examples, but I still don’t know how to define a correlation among a number of numeric attributes… I’m trying to generate a set that contain attributs correlated with each other (partly perfect colinearly). Is there a way to do that?


Hi Northern,

you might want to check out the paper here. He we explained the ideas behind those workflows and how they can be used.

Cheers, Iris

1 Like