Optimizing the distribution of distance between transactions - #Unplugged 21

Follow the process of optimizing a measure created in a previous unplugged to display the distribution of the distance between transactions in a column chart.
Many people who have seen the previous unplugged asked for this second episode to discover how the measure could be optimized.
How much faster can it go? Discover it in the video!


#unplugged

[martinbubenheimer]

Thus is so real life, I need to show this to my boss or customer next time when they are asking why does this take so long to implement! As you can see from the two videos the optimization takes more time then the initial solution. Thank you, Marco!

[pcorvalan]

Great!! I really appreciate watch a DAX master dealing with an optimization process ✨

[NirRobinson]

fabulous!!!

it took me about **3 minutes** to implement the optimize code in my system (I round the time 2 minute up :-)).

I instantly saw "new insight" and instantly add more measure same same but different.

your post are inspiring, easy to follow and worth any minutes I spend in your channel, book, website and course - and I spend a lot!!!!

thanks again!

Nir

[Maartenmarsje]

Great solution Marco! It's fun to see how you break down the challenge towards a solution. It really helps me when I'm working with my own complex formulas in DAX. Thank you so much for making this wonderful content.

[egorshiyanov]

That was one of the most awesome things I have ever seen! 10/10 !!

[stevenknight]

Was also very interesting to see how just changing the order the filter conditions applied made such a difference to performance. Thanks for another good video.

[kukfoongyew]

Truly an enjoyable time to understand your thought process.

[danielsiffredi]

I had to watch it a few times to get the intricacies off your thought process, but it was wonderful.
Thanks for sharing your thoughts and knowledge

[christianstork]

Beautifully explained! Thanks a lot. These Unplugged videos are great. BTW, "suspect" != "suspicion". :-)

[Lixia]

WoW!!! awesome result and I am impressed.

[jessamhan]

Due to the formidable reduction in query time it feels almost silly to write this, however, isn't the 'VAR previousDate = MAXX( _previousTxs, Sales[orderdate])' step superfluous?
The ' RankedDates' temporary table is based on VALUES(Sales[orderdate]) that always returns distinct values and hence the RANKX will never return the same rank twice.
Again it's a little silly to try to improve on a measure like the one you have just served us. I bet though that Alberto would find it all right :-)

[AHMEDALDAFAAE]

Thank you for this great video!!
Recently Microsoft deployed 2 functions: INDEX and OFFSET, have you tried out if they are good and optimized to do this job of nested iteration?
Thank you in advance