Good evening, dear colleagues!
I didn't find the same topic.
So, I imported xlsx into jmp, but my numeric values became nominal.
How to avoid this situation?
Hi @Leonard ,
There's a few different ways to do that, but here's some JSL I wrote a while back that fixes that. I think JMP 16 is supposed to have a setting within the import wizard that allows overriding commas.
Names Default to Here(1); //This replaces all commas by decimals dt = Current Data Table(); Col_Names = dt << Get Column Names( Nominal, String ); Nr = N Rows( dt ); For( l = 1, l <= N Items( Col_Names ), l++, One_col = Column( dt, Col_Names[l] ); For( k = 1, k <= Nr, k++, One_col[k] = Substitute( One_Col[k], ",", "." ) ); ); //This part turns the "nominal" columns back to "numeric" For( l = 1, l <= N Items( Col_Names ), l++, Column( dt, Col_Names[l] ) << Data Type( Numeric ) << Modeling Type( Continuous ) << Format( Best, 12 ) );
You can try going File > Preferences > Windows Specific and check the box at the top: Use JMP.... But, I actually haven't gotten much luck with that actually overriding the comma.
But, the JSL code should help.
Hope this helps!,
DS
"File > Preferences > Windows Specific and check the box at the top: Use JMP...."
@SDF1 I've unchecked this box Use JMP language and voila.