topic Re: Reading raw currency data with R$ prefix in SAS Enterprise Guide

Reading raw currency data with R$ prefix

luis_costa — Wed, 28 Oct 2020 18:11:59 GMT

I'm trying to read a raw data file with the R$ (brazilian's currency) prefix.

What I can do by now is to read as character, use compress() to only keep digits, and then input() with a 10.2 informat and it works fine.

But I want a more "elegant" way to read this file, like using a custom informat in input.

Is it possible to do something like that?

Re: Reading raw currency data with R$ prefix

data_null__ — Wed, 28 Oct 2020 19:18:07 GMT

You need the LOCALE system option and an NLS informat.

554  data _null_;
555     x = 'R$100,50';
556     y = input(x,NLMNY10.);
557     z = input(x,NLMNY10.2);
558     put (y z)(=NLMNY10.2);
559     put _all_;
560     run;

y=R$100,50 z=R$100,50
x=R$100,50 y=100.5 z=100.5 _ERROR_=0 _N_=1

Re: Reading raw currency data with R$ prefix

novinosrin — Wed, 28 Oct 2020 19:19:59 GMT

My oh my!! Guru @data_null__ How did you know that?

Re: Reading raw currency data with R$ prefix

ballardw — Wed, 28 Oct 2020 19:21:20 GMT

You might be able to use an NLMNLBRLw.d informat.

data example;
    x='R$123,456.78';
    y= input(x,NLMNLBRL32.);
run;

There is a corresponding form to display the values in the same form.

Re: Reading raw currency data with R$ prefix

data_null__ — Wed, 28 Oct 2020 19:33:13 GMT

@ballardw wrote:

You might be able to use an NLMNLBRLw.d informat.
data example;
    x='R$123,456.78';
    y= input(x,NLMNLBRL32.);
run;
There is a corresponding form to display the values in the same form.

I believe that @ballardw suggestion is best option because you don't have to change the LOCALE.

Re: Reading raw currency data with R$ prefix

data_null__ — Wed, 28 Oct 2020 19:34:54 GMT

@ballardw wrote:

You might be able to use an NLMNLBRLw.d informat.
data example;
    x='R$123,456.78';
    y= input(x,NLMNLBRL32.);
run;
There is a corresponding form to display the values in the same form.

I believe this is the better option because you don't have to fiddle with LOCALE.

Re: Reading raw currency data with R$ prefix

data_null__ — Wed, 28 Oct 2020 19:37:50 GMT

I didn't know I had to look it up.

Re: Reading raw currency data with R$ prefix

novinosrin — Wed, 28 Oct 2020 19:48:04 GMT

Much thanks @ballardw and @data_null__ . I learned something new. Great stuff!!!

Re: Reading raw currency data with R$ prefix

luis_costa — Wed, 28 Oct 2020 19:55:34 GMT

Thank you so much guys!

As @data_null__ said, both @data_null__ and @ballardw answers works perfectly but I will mark @ballardw answer as the solution because we don't have to use the locale option.

Thanks again!

Re: Reading raw currency data with R$ prefix

ballardw — Wed, 28 Oct 2020 20:02:14 GMT

@data_null__ wrote:

You need the LOCALE system option and an NLS informat.

554  data _null_;
555     x = 'R$100,50';
556     y = input(x,NLMNY10.);
557     z = input(x,NLMNY10.2);
558     put (y z)(=NLMNY10.2);
559     put _all_;
560     run;

y=R$100,50 z=R$100,50
x=R$100,50 y=100.5 z=100.5 _ERROR_=0 _N_=1

This was where I started as well. But I remembered that there were some locale specific formats I had encountered (while looking for something else) and thought there might be a Brazil version. So I searched the formats documentation for Brazil an lo and behold, there is one.