Expected delivery based on data analysis

[jilbobagins]

Howdi hi!

Ok I have a table that details delivered parts, main columns I’m interested in is with of columns [item] , [Qty Delivered], [transaction date]. So I was wondering through maybe a group function are you able to look at the statistical probability of your next [transaction date] of the item and it’s qty??

maybe it’s more a dax function but I do prefer queries

 

[pinarello]

So that you can check whether the result corresponds to your ideas, I have first of all developed a formula-based solution. To make the result easier to understand, I have generated the result over several single steps.

Depending on the volume of data, a formula solution would be sufficient here, or it could also be represented in a PQ query.

 

[jilbobagins]

It’s many thousands of lines. One supplier in particular is awful on delivery and trying to plan isndifficuly
based on their promises....so maybe a predicted delivery date and qty might be useful

 

[pinarello]

Regardless of whether it should be a solution with PQ, formulas, VBA or another tool in the end, your information is meager so far. Above all, you do not specify which factors should be taken into account in the calculation. I have shown you one, albeit quite simple, calculation method. But you haven't mentioned it with a single word.

Therefore, I think it makes sense if you first inform yourself about which planning methods there are and which of them could be the best for your situation. Then it can be checked if and how this can be applied best for your data model.

 

[jilbobagins]

Apologies if the response was short (problem with fat fingers on a phone!) I really do appreciate your efforts and yes they workbook/formula's certainly gives me what I need. Its just I do prefer PQ as its less flaky (if that makes sense)…. and as the data source is rather large it might be a better way to go?

Thanks again

 

[pinarello]

It would be nice if you could upload an example of your data table, where the column titles correspond to the original, but the data is anonymized. You can also tell us something about the data volume and the evaluation period. Because with large time periods, with the method shown by me in the Excel table, changes in consumption behavior are leveled very strongly.