Forum Discussion
MATRIX PREVENT DRILL DOWN
Hi Dellis81 ,
You can add a new level to your filtering something similar to this:
VAR NRSwitch =
SWITCH (
SwitchValue,
"100", -ExtNR,
"130", -FamilyNR,
"120", -RelatedPartyNR,
"140", -(FamilyNR+RelatedPartyNR),
"150", TotalNR,
BLANK ()
)
VAR CenterDescrOrig = ISFILTERED('CenterSetup (orig)'[Description])
RETURN
IF (HASONEFILTER(Table[Column]),
IF (
AND ( MAX ( NoteReceivables[Show Detail] ) = 0, CenterDescrOrig ),
blank(),
NRSwitch), NrSwitch)
The Table[Column] is the name that you use for the lowest detail of your matrix (cannot see what is the name of that column).
- Dellis816 years ago
Post Prodigy
Thank you for your response 🙂
The lowest value of the matrix is 'CenterSetup (orig)'[Description]. I inserted the proper
table[column] per your suggestion resulting in the below (partial) measure. Unfortunately - I still get a blank when I drill down.
VAR CenterDescrOrig = ISFILTERED ( 'CenterSetup (orig)'[Description] ) RETURN IF ( HASONEFILTER ( 'CenterSetup (orig)'[Description] ), IF ( AND ( MAX ( NoteReceivables[Show Detail] ) = 0, CenterDescrOrig ), BLANK (), NRSwitch ), NRSwitch )I have verified the 'CenterSetup (orig)'[Description] is the lowest level in matrix.
Again - thank you for your help, this forum is great because of people like you! - Dellis816 years ago
Post Prodigy
Good morning MFelix
I've been working on this issue - and put together a small dataset together w/example to provide. But, as my normal luck goes - your suggestion worked (in my small dataset). However, I dropped the same measure back into the larger dataset - resulting in same blank result. So I know your suggestion works, but unsure why it's not in the live file. The only difference I know of - is my larger datasets has many more tables and relationships. I have double/trippled checked the relationships I have in my small test file matches the same relationships in live file.Do you have any thoughts where I might go from here?
Thanks
- MFelix6 years ago
Super User
Hi Dellis81,
Measures are based on context so depending on the way you have the model setup and also the visualization you may need to adjust the measure.
The formula is based on the lowest level of the matrix visualization so you need to refer to that column.
Has you refer the formula works but it's on a different dataset so without any details on your current dataset is difficult to give you guidance.
- Dellis816 years ago
Post Prodigy
Hello MFelix
Ok, I've spent most of the day trying to prep a larger scale example, that actually mimics my live datafile. Follow link to download https://1drv.ms/u/s!AmBVCme14p7xlTYmh0oHTxJC1wrG?e=aJ6Csf
I have changed the example - but the concept is the same. The items hilighted in yellow represent subtotals - and as illustrated - when click on the + symbol - the values turn into blanks. I am hoping to retain the values. In my rresearch - I found an example of someone using measures for the isfiltered and hasonefilter expression - so I have built those into the formula - thinking flexibility in the calculation depending on which dimension is used in rows. The result before and after resulted in the blank.
After a lengthy set of code to derive the a prior year average column - below is the section of measure I think we are dealing with. If you wish to test without the filtered measures - delete the measure, and format the commented out DAX to fit your needs.
RETURN IF([MaxtrixHasOneFilter], //HASONEFILTER(COASetup[Report LineDescription]), IF ( AND ( MAX ( ISHeader[Show Detail] ) = 0, [MaxtrixIsFiltered]),//isfiltered(COASetup[Report LineDescription])), blank(), CYrorAvgColumn ),CYrorAvgColumn)Again -= thank you. Forcing myself to create a scaled down test model will be valuable longer term. I have challenges to post - so if you see something we aren't discussing - that is yet to come 🙂