How are derived features computed in the visual analysis module?
Hi,
I was wondering how the derived features (square, root, log) are actually computed. Is the function applied after rescaling if any ?
In particular for square root and logarithm which cannot take non negative values, what is the actual preprocessing used here? Is there a different rescaling used to generate the derived columns compared to the user-specified one?
Answers
-
CoreyS Dataiker Alumni, Dataiku DSS Core Designer, Dataiku DSS Core Concepts, Registered Posts: 1,150 ✭✭✭✭✭✭✭✭✭OptionsHi, @NicolasB
! Can you provide any further details on the thread to assist users in helping you find a solution (insert examples like DSS version etc.) Also, can you let us know if you’ve tried any fixes already?This should lead to a quicker response from the community. -
tgb417 Dataiku DSS Core Designer, Dataiku DSS & SQL, Dataiku DSS ML Practitioner, Dataiku DSS Core Concepts, Neuron 2020, Neuron, Registered, Dataiku Frontrunner Awards 2021 Finalist, Neuron 2021, Neuron 2022, Frontrunner 2022 Finalist, Frontrunner 2022 Winner, Dataiku Frontrunner Awards 2021 Participant, Frontrunner 2022 Participant, Neuron 2023 Posts: 1,595 NeuronOptionsIf you read python. You can get the code in the form of a Jupyter Notebook that the Visual ML users for a particular model you are creating.
Although this does not work for all model types. This may help you understand exactly what DSS is doing in your case.
This can be found in the Action Drop down on any Model Results.
-
Thanks for your answers, we are using dataiku version 8.
I did take a look at the jupyter notebook generated from the visual ML but it does not seem to replicate the piece of code needed to generate the extra features.
I tried looking at some of .json files in the corresponding model folders but I didn't find anything to help me understand how the extra features are actually computed.
Thanks,
Nicolas