The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends September 15. Request your voucher.
Hi,
I am trying to create a flag based in the user date selection from the slicer. I have a disconnected date table and then i have my main table. I have a date slicer and the date comes from the date table. When user selects the date from the slicer i need to use that date and create a flag measure. I cannot use calculated column to filter the rows as the date comes dynamically based on user selection. I created below flag measure but this giving me incorrect results. I am testing the measure with 12/01/2022 as the user selection date. If i use the same logic in the backend snowflake table as where clause i am getting 66841 rows but if use the same logic as measure in power bi i am getting way too many rows. Please check the below measure logic,
[User_Selected_Date] = 12/01/2022 (This comes from the slicer selection)
Solved! Go to Solution.
I think the issue might be the ALL, which you don't include in the COUNTROWS version of the measure. Removing that might fix the problem.
I would point out that it is best practice not to filter entire tables but only to filter the columns you need, so you could use
Test Sumx =
VAR SelectedDate = [Selected_Date]
RETURN
SUMX (
FILTER (
SELECTCOLUMNS (
VW_CONS_PC_AR_LOAD,
VW_CONS_PC_AR_LOAD[ac_doc_status],
VW_CONS_PC_AR_LOAD[pstng_date],
VW_CONS_PC_AR_LOAD[PC_DBCR_GC],
VW_CONS_PC_AR_LOAD[CLEAR_DATE]
),
( VW_CONS_PC_AR_LOAD[ac_doc_status] = "O"
&& VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate
&& VW_CONS_PC_AR_LOAD[PC_DBCR_GC] <> 0 )
|| ( VW_CONS_PC_AR_LOAD[ac_doc_status] = "C"
&& VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate
&& VW_CONS_PC_AR_LOAD[CLEAR_DATE] >= SelectedDate
&& VW_CONS_PC_AR_LOAD[PC_DBCR_GC] <> 0 )
),
1
)
Finally, your Snowflake query has a GROUP BY clause, which neither of the above measures does, so that will return 1 row per combination of posting date and clearing date, whereas the DAX could return multiple rows.
Hi @vivek_babu
I wanted to check if you had the opportunity to review the information provided. Please feel free to contact us if you have any further questions.
Thank you.
Use Performance Analyzer to get the query for the visual and run it in DAX Query View, or DAX Studio. That will show you where the Flag measure is being evaluated. Sometimes it can get tricky when you use a measure as a filter when the measure is not in the visual itself.
Also, does the visual contain a column which would uniquely identify a row in your data, or is the data being grouped? If the data is being grouped then MAX will return the max value for the entire group, not an individual row.
Hi @johnt75
The countrows logic is working and i am getting the correct rows based on the user date selection. But i want the filteredrows in a flag value like 1 and i will use this flag in my table visual to further filter the data. This is the countrows measure which works,
I think the issue might be the ALL, which you don't include in the COUNTROWS version of the measure. Removing that might fix the problem.
I would point out that it is best practice not to filter entire tables but only to filter the columns you need, so you could use
Test Sumx =
VAR SelectedDate = [Selected_Date]
RETURN
SUMX (
FILTER (
SELECTCOLUMNS (
VW_CONS_PC_AR_LOAD,
VW_CONS_PC_AR_LOAD[ac_doc_status],
VW_CONS_PC_AR_LOAD[pstng_date],
VW_CONS_PC_AR_LOAD[PC_DBCR_GC],
VW_CONS_PC_AR_LOAD[CLEAR_DATE]
),
( VW_CONS_PC_AR_LOAD[ac_doc_status] = "O"
&& VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate
&& VW_CONS_PC_AR_LOAD[PC_DBCR_GC] <> 0 )
|| ( VW_CONS_PC_AR_LOAD[ac_doc_status] = "C"
&& VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate
&& VW_CONS_PC_AR_LOAD[CLEAR_DATE] >= SelectedDate
&& VW_CONS_PC_AR_LOAD[PC_DBCR_GC] <> 0 )
),
1
)
Finally, your Snowflake query has a GROUP BY clause, which neither of the above measures does, so that will return 1 row per combination of posting date and clearing date, whereas the DAX could return multiple rows.
Hi @johnt75
This measure worked,
Hi @johnt75
Thanks for the response
This is the snowflake query,
SELECT pstng_date, clear_date
FROM DP_FDW.VW_CONS_PC_AR_LOAD
WHERE
(
ac_doc_status = 'O' AND pstng_date <= '2022-12-01'
)
OR (
ac_doc_status = 'C' AND pstng_date <= '2022-12-01' AND CLEAR_DATE >= '2022-12-01'
)
GROUP BY pstng_date, clear_date
ORDER BY pstng_date
I am getting 13202 rows as the output. Only difference here the date is hardcoded but in Power BI it should be dynamic and coming from user selection. The above measure logic is not working i am getting less rows as the ouput(2123 rows only) and some dates are missing as compared to snowflake result
Regards
Hi @vivek_babu
You should use FILTER to evaulate each row of your main table and apply the logic correctly. Try this:
Test Flag =
VAR SelectedDate = [User_Selected_Date]
RETURN
CALCULATE(
COUNTROWS(
FILTER(
VW_CONS_PC_AR_LOAD,
NOT ISBLANK(SelectedDate) &&
(
(VW_CONS_PC_AR_LOAD[ac_doc_status] = "0" && VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate) ||
(VW_CONS_PC_AR_LOAD[ac_doc_status] = "C" && VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate && VW_CONS_PC_AR_LOAD[clear_date] >= SelectedDate)
)
)
)
)
I hope this helps, please give a thumbs up and mark as solved if it does, thanks!
Hi @wardy912 ,
The countrows logic is working and i am getting the correct rows based on the user date selection. But i want the filteredrows in a flag value like 1 and i will use this flag in my table visual to further filter the data. This is the countrows measure which works,
Hi @wardy912
Thansk for the response,
This measure logic is not working. This is the snowflake query,
SELECT pstng_date, clear_date
FROM DP_FDW.VW_CONS_PC_AR_LOAD
WHERE
(
ac_doc_status = 'O' AND pstng_date <= '2022-12-01'
)
OR (
ac_doc_status = 'C' AND pstng_date <= '2022-12-01' AND CLEAR_DATE >= '2022-12-01'
)
GROUP BY pstng_date, clear_date
ORDER BY pstng_date
I am getting 13202 rows as the output. Only difference here the date is hardcoded but in Power BI it should be dynamic and coming from user selection. The above measure logic is not working i am getting less rows as the ouput(2123 rows only) and some dates are missing as compared to snowflake result
Regards
Thanks for the quick response,
I cannot directly access the table columns inside the measure. I am getting below error,
A single value for column 'ac_doc_status' in table 'VW_CONS_PC_AR_LOAD' cannot be determined. This can happen when a measure formula refers to a column that contains many values without specifying an aggregation such as min, max, count, or sum to get a single result.
Because of this reason i used MAX aggregate function in the measure but that is not working properly
Regards
Got it
SUMX would make this iterative.
For Measure try to use SUMX to make the formula iterative by row. If this doesnt help as well, it would be easier if sample data is provided.
I believe in filter context, you wouldn't have to use MAX, let me know if you get any error for this.
Test Flag Measure =
VAR SelectedDate = SELECTEDVALUE('DateTable'[Date])
RETURN
IF(
ISBLANK(SelectedDate),
BLANK(),
SUMX(
FILTER(
'VW_CONS_PC_AR_LOAD',
(
('VW_CONS_PC_AR_LOAD'[ac_doc_status] = "0" && 'VW_CONS_PC_AR_LOAD'[pstng_date] <= SelectedDate)
|| ('VW_CONS_PC_AR_LOAD'[ac_doc_status] = "C" && 'VW_CONS_PC_AR_LOAD'[pstng_date] <= SelectedDate && 'VW_CONS_PC_AR_LOAD'[clear_date] >= SelectedDate)
)
),
1
)
)
This measure logic is not working. This is the snowflake query,
SELECT pstng_date, clear_date
FROM DP_FDW.VW_CONS_PC_AR_LOAD
WHERE
(
ac_doc_status = 'O' AND pstng_date <= '2022-12-01'
)
OR (
ac_doc_status = 'C' AND pstng_date <= '2022-12-01' AND CLEAR_DATE >= '2022-12-01'
)
GROUP BY pstng_date, clear_date
ORDER BY pstng_date
I am getting 13202 rows as the output. Only difference here the date is hardcoded but in Power BI it should be dynamic and coming from user selection. The above measure logic is not working i am getting less rows as the ouput(2123 rows only) and some dates are missing as compared to snowflake result
Regards
How about using ALL function to consider the entire table. Sorry about going back and forth, this is because its a little difficult without the pbix to know if this would solve the issue.
TestFlagCount =
VAR SelectedDate = SELECTEDVALUE('DateTable'[Date])
RETURN
IF(
ISBLANK(SelectedDate),
0,
CALCULATE(
DISTINCTCOUNT('VW_CONS_PC_AR_LOAD'[pstng_date]),
FILTER(
ALL('VW_CONS_PC_AR_LOAD'),
(
('VW_CONS_PC_AR_LOAD'[ac_doc_status] = "O" && 'VW_CONS_PC_AR_LOAD'[pstng_date] <= SelectedDate)
||
('VW_CONS_PC_AR_LOAD'[ac_doc_status] = "C" && 'VW_CONS_PC_AR_LOAD'[pstng_date] <= SelectedDate && 'VW_CONS_PC_AR_LOAD'[clear_date] >= SelectedDate)
)
)
)
)
Or CountROWS
Test Flag =
VAR SelectedDate = SELECTEDVALUE('DateTable'[Date])
RETURN
IF(
ISBLANK(SelectedDate),
BLANK(),
CALCULATE(
COUNTROWS(
FILTER(
'VW_CONS_PC_AR_LOAD', --Incase none are working try adding ALL here as well
(
'VW_CONS_PC_AR_LOAD'[ac_doc_status] = "0"
&& 'VW_CONS_PC_AR_LOAD'[pstng_date] <= SelectedDate
)
|| (
'VW_CONS_PC_AR_LOAD'[ac_doc_status] = "C"
&& 'VW_CONS_PC_AR_LOAD'[pstng_date] <= SelectedDate
&& 'VW_CONS_PC_AR_LOAD'[clear_date] >= SelectedDate
)
)
)
)
)
Would also be helpful to know what error or issue is faced when trying these and any further input on the behavior
Thanks for the respone
The countrows logic is working and i am getting the correct rows based on the user date selection. But i want the filteredrows in a flag value like 1 and i will use this flag in my table visual to further filter the data. This is the countrows measure which works,
Hi @vivek_babu
Could you check if this works, the MAX might be getting one particular value from the column without proper row context.
Test Flag =
VAR SelectedDate = SELECTEDVALUE('DateTable'[Date]) --Basically trying to get the date the user has selected from slicer selection (Use the field used in the slicer here)
RETURN
IF(
NOT ISBLANK(SelectedDate) &&
(
(VW_CONS_PC_AR_LOAD[ac_doc_status] = "0" && VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate) ||
(VW_CONS_PC_AR_LOAD[ac_doc_status] = "C" && VW_CONS_PC_AR_LOAD[pstng_date] <= SelectedDate && VW_CONS_PC_AR_LOAD[clear_date] >= SelectedDate)
),
1,
0
)
Also give a try using selectedvalue in place of MAX just incase its appearing appropriately in each row.
Please provide further context if this doesn't help