I've got a table of maintenance activities and I'm struggling to figure out how to use DATEDIFF to calculate the number of days between the grass cuttings grouped by parks. Here's a sample of the data
Ideally, what I'd like to do is have a fourth column that says "Days since last cut". I've tried to modify code I've seen using DATEDIFF and some sort of placeholder VAR, but they all seem to assume that you've only got one "Park" value, so it gives me the number of days since ANY park was cut. Any help would be appreciated.
Here's what I came up with. There's a few steps, but not to painful.
Step 1:
- Get the data into Power Query.
- Figure out a way to remove exact duplicates. Meaning same Park, Activity, and Completed Date/Time
- Create a copy of the Completed column, and change to data type to Whole Number
Step 2:
- Load that into the data model
- Add an "Index" column so we know what the previous date was:
Index = VAR CurrentPark= 'Calculate Dates Between'[Park] Var CurrentAct= 'Calculate Dates Between'[Activity] VAR CurrentCompleted= 'Calculate Dates Between'[Completed] RETURN CALCULATE( COUNTROWS( FILTER( ALL ( 'Calculate Dates Between' ) , CurrentPark = 'Calculate Dates Between'[Park] && CurrentAct = 'Calculate Dates Between'[Activity] && CurrentCompleted >= 'Calculate Dates Between'[Completed] ) ) ) - Now we can write a measure since we have the data we need in the correct format:
Days Since Last Cut = IF ( NOT ( HASONEVALUE('Calculate Dates Between'[Index])), BLANK(), /* this will put blank if there is more than 1 index (i.e. subtotals/totals*/ IF( MAX('Calculate Dates Between'[Index]) <>1, /*do not want a value on the 1st date, so will put "First cut"*/ MAX( 'Calculate Dates Between'[Completed - Whole Number])- /*Current Whole Date in the current filter context*/ CALCULATE( MAX( 'Calculate Dates Between'[Completed - Whole Number]), FILTER( ALL( 'Calculate Dates Between'), 'Calculate Dates Between'[Index] = MAX('Calculate Dates Between'[Index])-1 ), /*want to filter that whole number column we created by taking the current index and going back one */ VALUES('Calculate Dates Between'[Park]) /*need to keep the filter on Park in play*/ ), "First Cut" /*Value for 1st date, could be anything )Seems like a lot (and it is!) but when broken down it's not so bad.....
Here's the final output:
Hopefully that makes sense, but fire away any questions
Here's what I came up with. There's a few steps, but not to painful.
Step 1:
- Get the data into Power Query.
- Figure out a way to remove exact duplicates. Meaning same Park, Activity, and Completed Date/Time
- Create a copy of the Completed column, and change to data type to Whole Number
Step 2:
- Load that into the data model
- Add an "Index" column so we know what the previous date was:
Index = VAR CurrentPark= 'Calculate Dates Between'[Park] Var CurrentAct= 'Calculate Dates Between'[Activity] VAR CurrentCompleted= 'Calculate Dates Between'[Completed] RETURN CALCULATE( COUNTROWS( FILTER( ALL ( 'Calculate Dates Between' ) , CurrentPark = 'Calculate Dates Between'[Park] && CurrentAct = 'Calculate Dates Between'[Activity] && CurrentCompleted >= 'Calculate Dates Between'[Completed] ) ) ) - Now we can write a measure since we have the data we need in the correct format:
Days Since Last Cut = IF ( NOT ( HASONEVALUE('Calculate Dates Between'[Index])), BLANK(), /* this will put blank if there is more than 1 index (i.e. subtotals/totals*/ IF( MAX('Calculate Dates Between'[Index]) <>1, /*do not want a value on the 1st date, so will put "First cut"*/ MAX( 'Calculate Dates Between'[Completed - Whole Number])- /*Current Whole Date in the current filter context*/ CALCULATE( MAX( 'Calculate Dates Between'[Completed - Whole Number]), FILTER( ALL( 'Calculate Dates Between'), 'Calculate Dates Between'[Index] = MAX('Calculate Dates Between'[Index])-1 ), /*want to filter that whole number column we created by taking the current index and going back one */ VALUES('Calculate Dates Between'[Park]) /*need to keep the filter on Park in play*/ ), "First Cut" /*Value for 1st date, could be anything )Seems like a lot (and it is!) but when broken down it's not so bad.....
Here's the final output:
Hopefully that makes sense, but fire away any questions
Thank you! Oh, man, I think I'm so close to being there. My column names are slightly different, but here's my code for the first part:
Index =
VAR CurrentPark= 'Maintenance'[Park Name]
Var CurrentAct= 'Maintenance'[Maintenance Task]
VAR CurrentCompleted= 'Maintenance'[Finish Time]
RETURN
CALCULATE(
COUNTROWS(
FILTER( ALL ( 'Maintenance' ) ,
CurrentPark = 'Maintenance'[Park Name]
&& CurrentAct = 'Maintenance'[Maintenance Task]
&& CurrentCompleted >= 'Maintenance'[Finish Time]
)
)
)It seems to work, but I get some weird index numbers when I create a table to check.
I'm assuming I'm messing something up. Additionally, I'm not sure how to de-dupe, but (in theory) the data shouldn't have dupes because it's coming from an ESRI app that doesn't allow for duplicates. That said, when I try to run a "Delete Duplicates" in Power Query, it's obviously looking for a true "all columns match" duplicate and I can't seem to find a way to edit that.
Thoughts?
There are definitely dupes in the raw data:
that's why the Index gets messed up there. Because it should be 6, but since there are two records, it adds them together to get 12 and then things get wonky.
To remove the dupes in PQ, I mergeed those three columns, and then used the remove duplicates in that column. Then you can actually delete that column out and everything will still work.
Probably could use something like distinctcount, but I'd prefer to clean up the data in PQ
Ok. So, now the Index works fine and the calculation seems to be working fine, but it's coming up with weird negative numbers. Disclaimer, in the original data set, there were situations where the field techs would hit the button twice, so we'd have multiple finish times on one day. But, I made a duplicate of the finish time column as a date only and used it in the dedupe field. So, the index looks right, but it gives me weirdness. See below:
Any ideas?
The duplicates need to be removed, which I'm not sure is happening on your end. I do not see those negative numbers:
I attached the pbix file below. Take a look at the query in Power Query to see what I did.
The index calculated column takes uses the Park, Activity, and Date Completed to as the "filter" to count, so adding in new activities shouldnt matter.
I added in a generic "Activity 2" and still get the same result:
if you want, share the pbix file. Not sure if you can but could be helpful. Unless you are good?
Helpful resources
Microsoft Fabric Learn Together
Covering the world! 9:00-10:30 AM Sydney, 4:00-5:30 PM CET (Paris/Berlin), 7:00-8:30 PM Mexico City
Power BI Monthly Update - April 2024
Check out the April 2024 Power BI update to learn about new features.
Fabric Community Update - April 2024
Find out what's new and trending in the Fabric Community.
