Running totals for ranked categories without using date?

[pborah]

Hi all,

I have 8 categories "Sales over 5000", "Sales over 10,000", ..... "Sales over 40,000". Clearly these categories are ranked but there is nothing in the dataset to tell PowerBI so. That is the first problem. Next I want to create a cumulative frequency or a running total table for this information sorted by the ranked categories in ascending order. I have read most of the cumulative freq/running total articles but it seems they all have something to do with date whereas I have no need for using date in my problem to the best of my understanding. Can anyone please explain how to go about it in such a scenario? Thank you.

[lbendlin]

First you need to decided if your buckets can be defined as a calculated column (not impacted by user filter choices) or as a measure (impacted by user filter choices)

And second, yes, you DO need a calendar table. How else are you planning to do the cumulative calculation?

[pborah]

Thank you @lbendlin  for your response. My buckets are in fact calculated columns. I'm looking to create a basic cumulative frequency (count or percentage does not matter) table that they teach you in elementary stats class such as this one -

Class	Frequency	Cumulative Frequency
0-5	2	2
5-10	6	8
10-15	13	21
15-20	17	38
20-25	11	49
25-30	4	53
30-35	2	55
	55

Can you please explain why I need a date column to have this table?? Thanks.

[pborah · Accepted solution]

After a lot of searching, here's what I went with -

First I created a measure called -

Transaction Count = CALCULATE(COUNT(Query1[Transaction Number]))

Next I created another measure -

TN Cumulative Count = CALCULATE([Transaction Count],FILTER(ALL(Query1[Sales Class]),Query1[Sales Class]>=MAX(Query1[Sales Class])))

 Now I select a table visual, and drop Sales Class, Count of Sales, and TN Cumulative Count in Values and voila!