How to obtain mean values for the observations in a dataset

[Banke]

Hello everyone, 

please i have to following dataset. The THCIC_ID variable is the unique ID of each hospital in PROVIDER_NAME variable column. Each observation per hospital is the inidividual patient that was admitted in that hospital. e.g austin state hospital, the first hospital, have 8 admissions. I am asked to provide the mean number of admissions across all hospitals. do i have to create a count ? i dont know how to go about it. There are 361 hospitals and over 9000 observations in the dataset.

Thank you

data set

[PaigeMiller]

PROC FREQ will determine the actual number of admissions per hospital.

To get the mean number of admissions, you would run PROC MEANS on the output of PROC FREQ.

[Banke]

Thank you. Can you please show me the syntax? 

is it something like

PROC FREQ DATA = WEEK6.THCICS;

TABLES PROVIDER_NAME;

OUTPUT OUT = MEANS MIN MAX;

RUN;

?

[PaigeMiller · Accepted solution]

proc freq data=week6.thcics;
    tables provider_name/out=counts noprint;
run;
proc means data=counts;
    var count;
run;

[Banke]

It worked, thank you

[Kurt_Bremser]

You ca run two nested SQL SELECTs.

proc sql;
select mean (count)
from (select hosp_id, count(*) as count from have group by hosp_id);
quit,;

In the future, please supply data in usable form, in a data  step with datalines, not as pictures, so we have something to test against.

[Banke]

Thanks, i do not know how to go about it. t hats why i dont have a data step, i dont know about proc sql yet, can you please explain to me:?