Core Designer Course| Request for feedback
Hi!
I had a dataset with a column for country, a column with 5 different years x country and a column with some statistics.
I wanted to:
1) Rank each country for each year on that statistic
2) Compute difference between the stat of yeach year compared to the previous year
What I did:
1) Pivoted the df, so that I had one row per country and five columns for the value of the statistic for each year
2) I added 5 different Sort Recipe to the flow to add 5 columns with the rank of the country for each year (and here it comes the first feedback request, was there a faster way to obtain the rank on different columns?)
3) Added a Prepare Recipe with 4 formulas to compute 4 columns with change on that statistic from an year to the previous one (again, was there a faster way? At the beginning I tried to obtain this using row LagDiff in a Window Recipe before pivoting the data but then I got stuck on the ranking part) and canceled the columns with the stats.
Thank you, any feedback is appreciated!
Answers
-
Michael Grayson Administrator, Dataiker, Alpha Tester, Dataiku DSS Core Designer, Community Team Posts: 299 Administrator
Hi Beatrice, I've moved your post to the General Discussion board since I think you are more likely to find relevant help here.
Thanks!