Rolling average with working days in DAX

This article shows how to compute a rolling average (useful to smooth out charts) taking into account only the working days.
Article and download: sql.bi/757611?...
How to learn DAX: www.sqlbi.com/...
The definitive guide to DAX: www.sqlbi.com/...

[@Tom-kp2lv]

Thanks!! So, import and a calc col is and available option in my scenario. Is the "add columns"/measure version slower or faster than using a calculated column?

[@mnowako]

That’s a very interesting business case. Both examples are great and the explanation is second to none. Thank you and please make more such examples. 👏

[@readistreet9383]

Hi, great video as allways, it would help me a lot. Just one thing, the first values of the moving average are not correct, till it reach the 30 days of data. for example, in the first day youre dividing the first value by 30 wich is not correct. Maybe a better way is to use in the VAR NumberOfDays the actual number of days in the range min/max workingday instead of just 30

[@matthewdufty606]

The KZbin algorithm literally read my mind. I wanted to look at rolling averages this morning and this popped up in my feed. How bizarre!

[@PJJACOB360]

You are the best

[@matthiask4602]

Alberto, thank you so much for videos like this. You are the best teacher of data analytics (esp. DAX of course) I ever experienced.

[@graemegourlay2850]

Exactly what I needed to do for a piece of analysis. I would never have figured this out myself. Big thank you for this tutorial.

[@FitofnietOfficial]

Thanks for the great content. I was wondering why did you add the less "

[@naylord5]

This video is excellent master! Thank you for sharing your knowledge and experience with the community. Can I ask you a question? I have a statement with a list of payments, each movement is associated to a document number, the problem is that the payments can be cancelled N amount of times and reapplied, either to the same doc or to another one. I would like to use Power Query to keep only the last odd (non) occurrence of the payments, but if the number of payments is even, discard them all. That is, if doc 1 had 1 payment, 1 cancellation and 1 payment again, only keep the third transaction, but if doc 2 had a payment and it was cancelled, but it was not paid again, discard all the payments of doc 2. Do you think you could help me? Thanks in advance for your time.

[@MS-im2zy]

Thank you! Doing this in a measure is exactly what I have been working on...and was stuck

[@kesavaraju9652]

In live connection also we can create a calculated column, By simply allowing a local model. But You showed an excellent way instead.

[@Anthony_Lecoq]

Crazy amazing good job Alberto as always. I love when you demonstrate how to solve an issue when starting from scratch. 👍

[@datasapien]

Amazing content Alberto! , the live calculated column part was pure wizardy.

[@brizotfrizot]

What is the relationship between your date table and sales table?

[@justFast1k]

Amazing! As usual Thank you, Alberto!

[@Bharath_PBI]

That's best as always! You mentioned IF() should be avoided in calculated column as best practice (optimization), do you have any article on this or any reference link?

[@SQLBI]

www.sqlbi.com/articles/calculated-columns-and-measures-in-dax/

[@AgulloBernat]

I love these vídeos!! Keep them coming!

[@PeterKontogeorgis]

You always make things look so easy! Looking forward to your training in Melbourne in October!

[@SQLBI]

Thank you Peter! Just one note: Marco will be teaching in Melbourne this year, not Alberto.

[@irfandaud7380]

DAX needs to be organized and should be in sequence ... Now i know why people say DAX is difficult

[@andreikilla]

What if you have multiple products in same table

[@SQLBI]

Use the product name as a grouping condition so you merge them.

[@firdosesyed8981]

Hlo sir I'm a beginner how much time it takes to learn the advanced level if SQL

[@sbn4862]

To learn DAX one must training every day, it's single way! What if we accept one filter in Calculate as WorkingDay="True"?

[@SQLBI]

The issue is how to count the number of working days so that they are always the same number in each interval. This is the purpose of the calculation made in the video.

[@frankgovers8420]

Should the not function have brackets?

[@AgulloBernat]

Question: the removefilters on date you add on 23:28 was not necessary in the calculated column version of the code?

[@SQLBI]

Not necessary there.