Plotting average time by no. of weeks, but counting only weeks with more than 1 data point.

[rainynights]

Essentially, what I'm trying to do is plotting the average marathon time (y-axis), by the number of weeks a person has been running for, while trying to exclude data from weeks there is only 1 person.

This picture provides a great graphical illustration of it: https://i.gyazo.com/08544e711adb796da5354e50dd7806ca.png

Essentially, I'm trying to plot avg time to finish a run for each runner that has been running for 1 week, for 2 weeks, for 3 weeks, etc.

However, one thing I'm also noticing is that weeks in which we only have data for 1 person are significantly skewing the results.

For example, in the picture, we see that the avg running time for the 47 weeks mark is really high, because only 1 person has been running for 48 weeks, and because their run-time that day was really slow, they skewed that figure really high.

What I'm trying to do is amend the measure (which at the moment is just taking all averages) to only include weeks in which more than one person ran, but I'm not even sure where to start!

I think it would also be amazingly helpful if this could be added as a what-if parameter (i.e. so the dashboard consumer can set the minimum threshold of runners needed as they wish).

[mwegener]

Hi @rainynights ,

has your question been answered?

If I answered your question, please mark my post as solution, this will also help others.

Please give Kudos for support.

[mwegener · Accepted solution]

Hi @rainynights ,

SUMMARIZE your data

https://docs.microsoft.com/en-us/dax/summarize-function-dax

FILTER your data

https://docs.microsoft.com/en-us/dax/filter-function-dax

get the AVERAGEX of your data

https://docs.microsoft.com/en-us/dax/averagex-function-dax

If I answered your question, please mark my post as solution, this will also help others.

Please give Kudos for support.