How to pull distinct counts from two different tables at once at align rows - sql

Question

I have two separate tables: emr and treatment. Each table has a userID column and a provider column. Currently, I'm doing a simple pull to count the number of distinct userIDs that appear in the emr table like this:
SELECT distinct vender, count (distinct userID) AS EMR_Patients
from emr
group by 1
This gets me the following output:
vender | EMR_Patients
+++++++++++++++++++++
a 10,000
b 5,000
c 37,500
However, I want to include the number of userIDs that also appear in the treatment table so I can see how many userID's that have an emr record and also have a treatment of interest. The output I'm trying to get is:
vender | EMR_Patients| Treatment_Patients
+++++++++++++++++++++++++++++++++++++++++
a 10,000 4,000
b 5,000 3,000
c 37,500 9,000
I tried using a union:
SELECT distinct vender, count (distinct userID) AS EMR_Patients
FROM emr
GROUP BY 1
UNION ALL
(SELECT distinct vender, count (distinct userID) AS Treatment_Patients
FROM treatment
GROUP BY 1)
But this doesn't work correctly. Is there a way to do this as a union, or should I left join the two tables together beforehand? Or maybe there's a cleaner way than either of these options?

Answer 1

Use JOIN:
SELECT e.vendor, e.EMR_Patients, t.Treatment_Patients
FROM (SELECT vendor, count(distinct userID) AS EMR_Patients
FROM emr
GROUP BY 1
) e
(SELECT vendor, count(distinct userID) AS Treatment_Patients
FROM treatment
GROUP BY 1
) t
ON e.vendor = t.vendor;
(I adjusted the spelling of "vendor".)
This will only include vendors in both tables. If you want vendors that are missing from one of the tables, you need an outer join of some sort. Your question is not clear on that.