topic Re: subsetting with the use of where and upcase function in SAS Programming

subsetting with the use of where and upcase function

Karim18 — Wed, 14 Jul 2021 14:10:07 GMT

data cars_subset;
set sashelp.cars;
where upcase(drivetrain)="FRONT";
RUN;

data cars_subset;
set sashelp.cars;
where drivetrain=upcase("front");
RUN;

first one produces 226 obs and 15 columns

and second one produces 0 obs and 15 columns

why it produces two different results ??

and please brief me about how upcase function works like this situation?

Re: subsetting with the use of where and upcase function

ErikLund_Jensen — Wed, 14 Jul 2021 14:16:13 GMT

Hi @Karim18

The value of the variable DRIVETRAIN in the CLASS data set is "Front".

If you upcase the variable value the result is "FRONT", so the first comparison is true.

But in the second comparison the original variable value "Front" is compared to upcase("front"), also ti "FRONT", which is false.

Re: subsetting with the use of where and upcase function

Reeza — Wed, 14 Jul 2021 14:19:09 GMT

Your first set of code:

where upcase(drivetrain)="FRONT";

Note that DRIVETRAIN and FRONT are both upcased in this filter.

Your second set of code:

where drivetrain=upcase("front");

In this one FRONT is upcased but we don't know if DRIVETRAIN is, unless it was done previously. Given the results, it seems like it's not upcased so both sides of your equal signs are not the same. You could be comparing "Front" to "FRONT" which will not match.

@Karim18 wrote:

data cars_subset;
set sashelp.cars;
where upcase(drivetrain)="FRONT";
RUN;

data cars_subset;
set sashelp.cars;
where drivetrain=upcase("front");
RUN;

first one produces 226 obs and 15 columns

and second one produces 0 obs and 15 columns

why it produces two different results ??

and please brief me about how upcase function works like this situation?

Re: subsetting with the use of where and upcase function

Karim18 — Wed, 14 Jul 2021 15:09:44 GMT

sir upcase function manipulates data in execution phase and where statemt is compiled time statement

then how come it changes to uppercase???

Re: subsetting with the use of where and upcase function

Reeza — Wed, 14 Jul 2021 15:27:51 GMT

Timing of when something happens doesn't matter in this particular situation, it has to do with the values not matching.