Forum Discussion
Chitemerere
6 years agoResponsive Resident
Extract values from string with DAX
I have an attribute (column) called "Dissolution" from a table in Power BI which is a string with values in the string. I want to extract the values as new attributes (columns). The sample column is...
- 6 years ago
Chitemerere - That's pretty ugly but I believe these will work:
Column 1 = VAR __Pos = SEARCH("%",[Dissolution],,-1) VAR __Col = IF(__Pos=-1,-1,SEARCH(":",[Dissolution],__Pos - 8,-1)) RETURN IF(__Pos = -1 || __Col = -1,BLANK(),MID([Dissolution],__Col+2, __Pos-__Col-1)) Column 2 = VAR __Pos1 = SEARCH("%",[Dissolution],,-1) VAR __Pos = SEARCH("%",[Dissolution],__Pos1 + 1,-1) VAR __Col = IF(__Pos=-1,-1,SEARCH(":",[Dissolution],__Pos - 8,-1)) RETURN IF(__Pos = -1 || __Col = -1,BLANK(),MID([Dissolution],__Col+2, __Pos-__Col-1)) Column 3 = VAR __Pos1 = SEARCH("%",[Dissolution],,-1) VAR __Pos2 = SEARCH("%",[Dissolution],__Pos1 + 1,-1) VAR __Pos = IF(__Pos2 = -1,-1,SEARCH("%",[Dissolution],__Pos2 + 1,-1)) VAR __Col = IF(__Pos=-1,-1,SEARCH(":",[Dissolution],__Pos - 8,-1)) RETURN IF(__Pos = -1 || __Col = -1,BLANK(),MID([Dissolution],__Col+2, __Pos-__Col-1))
Greg_Deckler
6 years agoCommunity Champion
Chitemerere - It's hard to say, it looks like it is inventing numbers that aren't actually in the row/column so no idea how that could be happening. You are implementing this as a column, correct? Not a measure?
Like your first row,
Complies at S1
mean (n=6):
Lamivudine: 99.8%
Zidovudine: 101.0%
I don't see where the calculation could possibly get 100.00% That does not even seem possible to me because that number isn't even in the text anywhere. I'm attaching my working PBIX. You want Table (2). It is based on your original sample data.
Chitemerere
6 years agoResponsive Resident
Many thanks, managed to resolve the problem.
Regards,
Chris