Cohort Analysis with Synthetic Dimension

Tue, 29 Apr 2014 05:02:41 GMT

Hi there,

I've been stuck trying to develop a cohort analysis for a client.

My problem is that the client needs very specific functionality when comparing cohorts.

For example, cohorts can be defined by the following:

Cohort 1: Cohort_Start_Month = Jan 2013, Region = Region A, Product = Product X

Cohort 2: Cohort_Start_Month = June 2012, Region = Region A, Product = Product Y

The difficulty here is that he wants a line chart to compare measures like sales amounts over time but using a common dimension with values, +1, +2, +3 etc. (so for Cohort 1, Jan 2013 = +1, Feb 2013 = +2 etc. and for Cohort 2, June 2012 = +1, Aug 2012 = +2, etc.) to easily compare the two cohorts over time. My thinking was that I need to convert each month from each cohort into these dimension values and then match them somehow. I've tried matching a valueloop() function to a variable that calculates these values but no luck.

Any ideas would be greatly appreciated.

Cheers,

Rhys

Re: Cohort Analysis with Synthetic Dimension

jagan — Tue, 29 Apr 2014 07:25:52 GMT

Hi,

Can you attach some sample file?

Regards.

Jagan.

topic Cohort Analysis with Synthetic Dimension in QlikView

Cohort Analysis with Synthetic Dimension

Re: Cohort Analysis with Synthetic Dimension