Stupidly constructed data set..

[Adam Wysokinski]

Hi all,

I use (and love) Stata 13. I've got a data set (>67,000 observations), which is far from being perfectly constructed.

Here's how my data looks:

patient_ID examination result cytosis anemia

001 MCV 90 normal

001 HGB 14 no

002 MCV 80 micro

002 HGB 10 yes

I want to check how many patients with anemia of particular type (with cytosis normal, micro or macro) are in my sample.

I need a table which will combine cytosine and anemia. The command:

tab anemia cytosis 

returns:

no observations

(which was expected).

Any idea how to do it?

Regards,

Adam