0 votes
I am trying to reshape a dataset using pivot. My dataset structure is like this : ['column A', 'column B', 'index', 'label', 'value'].

I have preprocessed the data using a Python script : 1) sorting the index column I want to use as my pivot, and 2) reordering the columns so that my dataset is now : ['index', 'label', 'value', 'column A', 'column B'], since I noticed the pivot method would work inconsistently otherwise.

Now the visual pivot recipe gets the job done, but the python pivot method fails to do so : all other columns are removed, and I cannot find additional arguments to correct this.

Is this discrepancy between your processor and the Pandas method known to you ? The Pandas pivot method seems highly unpractical in its current state, what alternative would you recommend since I really want to be able to process my data programmatically ?

edit : using DSS 4.0.4, 2-week trial of the Entreprise Edition, working from my own machine.
Not sure I understand the issue here.

You could run the visual pivot and use its output for your code recipe?

Please log in or register to answer this question.

1,299 questions
1,327 answers
11,867 users

┬ęDataiku 2012-2018 - Privacy Policy