问题
I'm trying to use the RANKX formula rank the values of one column, but filtered for the value of a second column. In this example, col2 is a simple counter running in ascending value. I'm trying to find item_id's Rank value relative to the col1.
col1 col2
1001 8001
1001 8002
1002 8003
1002 8004
1002 8005
I'd like to figure out a col3 that would read:
col1 col2 col3
1001 8001 1
1001 8002 2
1002 8003 1
1002 8004 2
1002 8005 3
Because that would be the rank of col2 relative to col1.
回答1:
You don't need to use RANKX at all. See the section on EARLIER in the superb http://www.daxpatterns.com/. Add a new calculated column in your table:
=
COUNTROWS (
FILTER (
MyTable,
[col2] <= EARLIER ( [col2] )
&& [col1] = EARLIER ( [col1] )
)
)
If you do want to use RANKX you can adapt the formula as follows:
=
RANKX (
FILTER ( MyTable, [col1] >= EARLIER ( [col1] ) ),
[Col2],
,
1
)
来源:https://stackoverflow.com/questions/29857158/rank-values-of-one-column-by-filtering-on-2nd-column-in-dax