Unexpected behavior of not equal ={'*'}- in set analysis

[fashid]

Hi Experts,

I have two same set analysis expressions

1)    =  ((SUM({<SERVICE_LINE={'*'}-{'CEN','STG'}>}GP_TARGET)/100))

            /count({<SERVICE_LINE={'*'}-{'CEN','STG'},GP_TARGET={'*'}-{0}>}GP_TARGET_KEY)

2)  =((SUM({<SERVICE_LINE-={'CEN','STG'}>}GP_TARGET)/100))

        /count({<SERVICE_LINE-={'CEN','STG'},GP_TARGET={'*'}-{0}>}GP_TARGET_KEY)

The only difference between these two expressions is the way not equals to is used for service_line field.

I have a difference in behavior with these two expressions.

When i click on a particular service_line list box the second expression changes and gives the desired result according to the selections made but the first expression remains static and does not change .

Is this behavior correct and should i always use the ={'*'}-

Another question is if i want to compare a value to 0 i.e. GP_TARGET not equal to zero should i write the way i have written above or should i use GP_TARGER<>0

Regards,

Nadeem

[rwunderlich · Accepted solution]

In your first expression you have:

SERVICE_LINE={'*'}

Which means start with "select all values" -- which effectively ignores any selections you have made in SERVICE_LINE.

-Rob

http://masterssummit.com

http://robwunderlich.com

[jonathandienst]

Rob is correct - the first expression says select all and then exclude.

The second expression is shorthand for

     $-<...>

which means take the current selection and then exclude.

Hope that clears it up.

[fashid]

Thanks jonathan you last line "which means take the current selection and then exclude." made it more clear

Regards

Nadeem