Question

One column in my data is sample, another is category. Duplicate are allowed. Number of unique categories I currently have is 5. Here is a simplified example:

sample category other_columns
122    a
123    a
124    a
125    a
123    b
124    b
125    b
122    c
123    c
124    c
...    ...

I need to select only those samples that exist in all categories (122 is not in ‘b’ and 125 is not in ‘c’).

sample category
123    a
124    a
123    b
124    b
123    c
124    c

So, if I run

SELECT category, COUNT(DISTINCT sample, category)
FROM my_table
GROUP BY category

all counts should be the same.

Answer 1

This one should do the trick:

SELECT sample 
FROM my_table 
GROUP BY sample
HAVING count(DISTINCT category) >= (SELECT count(DISTINCT category) FROM my_table);