DAX for Power BI - Predicting Days Until a Goal (Linear Regression)

[@mateusalves8991]

Man! This was one of the best vídeos i ever watched here at youtube! quick, objective, functional and pretty usefull! thanks for share!

[@BIElite]

Wow thanks Mateus! I appreciate the compliment :)

[@rerangelt]

Well, i just starting a project with forecasting. And i was thinking how i am gonna aproach it. You came just in time. Very well explaind. Thanks !!

[@BIElite]

Nice! What are you planning on forecasting?

[@jorgeqveloso]

AMAZING! So helpful! Greetings from Brazil!

[@barttitulaerexcelbart9400]

Awsome video thank you. In Excel you have the solver, you could use this to "optimise" your best time slot in order to achieve the highest slope (best scenario) or achieving the highest r squared to get the most "reliable" regression line. I think it will be a matter of time that SLOPE and INTERCEPT will be available in Power BI.

[@BIElite]

That would be awesome if/when they implement this. Thanks for bringing that up! I can think of a lot of use cases where those built-in calculations would be very useful

[@poojithabathini7291]

That was a great video, If there is a chance could you please explain how should we implement a situation where we want to achieve x amount of sub in n days then how can we do it?

[@TheHitamnoah]

Thanks for the video. Really appreciate it.

[@BIElite]

No problem!

[@JawadAli-nv9mr]

That was great video. I will request if you can make a video on( find a previous month/year dates, when same amount of target sales are achieved in current month or year.) That will be great thanks

[@shailendrakumar08]

Hey, thanks for the video. Could you please add raw data set for practice?

[@BrettvanGelder]

Great video. Thanks. Any recommendations on how to get the projected final amount as at a specified date (instead of the days until a target amount)? The use case is to forecast what the actual spend on a project will be at the end of the fiscal year based on the regression line.

[@fulyaulutas4599]

this is really cool . thank you

[@BIElite]

No problem, Fulya!

[@BajkeTheSerb]

Awesome! Must try!

[@BIElite]

+Predrag Bajić definitely! Thanks for watching

[@nicholaskeogh1121]

fantastic video, thanks for walkin gme through it. Question, is there a way to extend your Linear Regression line out to where it hits your target, and/or having a card set up to show the date at which your target is hit? i imagine the latter could be achieved by adding the "days until target is met" to the latest date of your data maybe...

[@BIElite]

Hey Nicholas! Sorry for the crazy late response. I guess the only way I would know how to solve this would be to use the method shown around the 14:25 mark of the video to calculate the number of days until the goal is hit and add to Today or the latest date of your data like you said. In order to extend the line in the chart would probably be more difficult...

[@bwilliams060]

This is great, can I grab the file somewhere?

[@BIElite]

Hi Ben, I don't have this file any more unfortunately.

[@omkar1o]

Hi !!! you video is awesome. I have something going on in mind to predict When my server disk space will be going to be full? I did use the regression formula you have used. Which is giving a nice trend line. and I am able to see the future data as well. but when I tried to filter it with Future date create a similar card as you have created, it's not able to show the data. Please let me know if you can share some tips, how I can achieve to show just predicted data

[@milacombs4586]

Hi Parker, I am trying to filter the x-axis date range based off of measures that the users of the report can change (using a what if analysis). I don't want to use an external slicer that's connected to the time column. Should I be filtering inside the running total measure and the linear regression/slope measures? Just one or the other? The report user won't be seeing the line graph, just the result of the "days until goal" measure. Thanks!

[@miltinhocamo3754]

Thank you Parker... awesome video... Many topics in only 16:47 min. Do you know about Cohorts?...

[@BIElite]

Thanks Miltinho! No I haven't heard about cohorts. What is that?

[@miltinhocamo3754]

It's a statistical analysis of data by groups, and do it with dax is pretty difficult.

[@sadyaz64]

awesome thanks. Next machine learning with dax?

[@BIElite]

Haha that would be really impressive. I'll leave that for someone smarter than me

[@jazzista1967]

Parker. Great trick. Just a quick question: Is the goal table a disconnected slicer? I am trying to replicate your example in powerpivot in excel 2016

[@BIElite]

Yep, you're absolutely right. The goal table is really just used to create that slider and provide us with the Goal Measure. Then that Goal Measure value can be used in any other calculations we need without needing any relationships

[@jazzista1967]

what a clever trick.. my head is still spinning..

[@christianfcbmx]

hI!!! Thanks for the Video...is there a way to use other option instead of "what if parameter" ...I always have problems when I have to use numbers over 1000...if it is over the goal field doesnt recognize the exat number you put in the box ...for example you need to put 2116 and the box sitches to 2115 ... thanks!

[@BIElite]

Hey Christian, I know exactly the issue you’re talking about. It’s very weird that the numbers don’t always line up like they should. You can always use normal parameters but it won’t be quite as dynamic.