HI every one

[SonyTrinesh]

Hi Everyone,

I have a Data set ,it contains character variables .some of the records are missing my requirement is how many records are missing in each variable.i tried with Freq and means.Means and summary giving stats only for numeric variables but i my Data set is contains character variables. is their any alternate method to get the requirement

Example:

DATA _null_;

Input Customer_Id $ Gender $ location $;

Cards;

1001  M   XXXX

1002       YYYY

         M   ZZZZ

          F   

;

RUN;

My Requirement is :

column_name     Miss  Nmiss

custom_id          2          2

Gender               1         3

location              1          3

[PGStats]

There may be more elegant ways but it can be done with proc freq, if you grab the output table with ODS :

DATA have;
Input Customer_Id $ Gender $ location $;
Cards;
1001  M   XXXX
1002  .   YYYY
.     M   ZZZZ
.     F .  
;

proc freq data=have;
tables _character_ / missprint;
ods output onewayfreqs=owf;
run;

data missCounts;
retain miss;
set owf; by table notsorted;
if first.table then miss = frequency;
if last.table then do;
     nonMiss = cumFrequency;
     var = scan(table, 2);
     output;
     end;
keep var miss nonMiss;
run;

proc print data=missCounts noobs; run;

PG

[Amir]

Hi,

You might also find the other responses to your similar posts helpful:

they appear to be "missing" a response from the OP.

Generally speaking, it is good practice to post your question just once, as mentioned in:

Regards,

Amir.