How to estimate count for Pandas dataframe column values based on multiple conditions?

Given a pandas DataFrame df of the following format:-

StudyNo   Condition1      Condition2      ResultVal
1222          25              30              0.5
1245          30              45              0.12
1672          40              15              0.46
1222          25              30              0.95
1245          30              45              0.88
1245          30              45              0.23

I would like to calculate frequency/count of the number of ResultVal attributes existing in the dataset for a specific StudyNo under a given Condition1 and Condition2. There are no exactly duplicate rows in this dataset.

e.g. for StudyNo 1245 at Condition1 of 30 and Condition2 of 45, the count would be 3 (because 3 ResultVal attributes exist for this aspect). Also, I would like to print out a list of StudyNo, Condition1 and Condition2 in cases wherein < =1 ResultVal attributes exist (which in this example would be [1672,40,15].

Any help in this regard would be much appreciated.

>Solution :

You can groupby "StudyNo", "Condition1" and "Condition2" and transform count of "ResultVal" to create column "Frequency".

Then create a boolean mask for rows where "Frequency" is less than or equal to 1 and filter the rows that satisfy it.

cols = ['StudyNo','Condition1','Condition2']
df['Frequency'] = df.groupby(cols)['ResultVal'].transform('count')
appears_only_once = df.loc[df['Frequency'] <= 1, cols].to_numpy().tolist()

Output:

   StudyNo  Condition1  Condition2  ResultVal  Frequency
0     1222          25          30       0.50          2
1     1245          30          45       0.12          3
2     1672          40          15       0.46          1
3     1222          25          30       0.95          2
4     1245          30          45       0.88          3
5     1245          30          45       0.23          3

The list:

[[1672, 40, 15]]