IS THERE ANY FUNCTION TO REPLACE VALUE "0" TO Missing for entire dataset

[venkatard]

IS THERE ANY FUNCTION TO REPLACE VALUE "0" TO Missing for entire dataset

[DBailey]

not a function...but...

proc sql;

update ds1

set col1=null where col1=0;

quit;

[venkatard]

This will only replace for one variable. I need to replace for entire dataset

[Amir]

Hi,

Is this just for one column in a data set, if yes then a simple solution would be to use a data step including the following line of code:

if my_var=0 then my_var=.;

I did notice that you surrounded the zero value with quotes ("0"), so if it is a character variable then you could use:

if my_var="0" then my_var="";

Regards,

Amir.

[Amir]

Hi,

For a function you could conditionally use in a data step:

call missing(my_var);

and it wouldn't matter if the variable is numeric or character.

Regards,

Amir.

Message was edited by: Amir Malik - added data step reference.

[venkatard]

I want to replace for all the variables.

[esjackso]

If there are multiple variables in the dataset that need to be converted you could combine post with an array:

data new;

     drop i j;

     set old;

     array c(*) _char_; /*or all your char variables that need to be converted*/

     array n(*) _numeric_; /*or all your numeric variables that need to be converted*/

     do i = 1 to hbound(c); /*sets the defined char vars where 0 to blank*/

          if c(i) = "0" then c(i) = "";

     end;

     do j = 1 to hbound(n); /*sets the defined numeric variables where 0 to missing*/

          if n(j)=0 then n(j)=.;

     end;

run;

Hope this helps!

EJ