Freq group count

[Barkamih]

Hi all,

 I have a big dataset (cows data) and I would like to know how many herds in each county in each month.

my data looks like this;

data example;
  
datelines;
county    herd  month 
London    11    Jan
London    11    Jan
London    11    Jan
London    11    Jan
London    11    Jan
London    12    Fed
London    11    Fed
London    12    Fed
London    11    Fed
London    12    Fed
London    12    Fed
London    11    Mar
London    12    Mar
London    13    Mar
London    11    Mar
London    12    Mar
London    14    Mar

And what I need to count the frequency of herd in each month by county and the results should be like this;

county     month     number of herds 

London    Jan       1 Herd

London    Feb       2 herds

London     Mar      4 herds 

So, I need a code to count frequencies of herds for each month in each county.

What My code should look like?

best regards 

Ibrahim 

[novinosrin · Accepted solution]

data example;
input county $   herd  month $;
datAlines;
London    11    Jan
London    11    Jan
London    11    Jan
London    11    Jan
London    11    Jan
London    12    Fed
London    11    Fed
London    12    Fed
London    11    Fed
London    12    Fed
London    12    Fed
London    11    Mar
London    12    Mar
London    13    Mar
London    11    Mar
London    12    Mar
London    14    Mar
;

proc sql;
create table want as
select county,month, count (distinct herd) as num_of_herds
from example
group by county, month;
quit;

/*Or*/

proc sql;
create table want as
select county,month, catx(' ',count (distinct herd),'herds') as num_of_herds
from example
group by county, month;
quit;

[Barkamih]

 Thaaaaaaaaaaaaaaaaaaaaaaaaaaaanks million 

It's done 

working perfectly 

best wishes 

[ballardw]

Another approach is two proc freq calls:

proc freq data=example noprint;
tables county*herd*month/out=work.firstcount;
run;
proc freq data=work.firstcount;
tables county*month/list;
run;

However the sort order is based on text so Jan does not appear before Feb.