Applying a mask to a certain range in a pandas column

user_13370365 0
Viewed 62

I'm currently trying to apply a mask to a column on a dataframe, in order to gain the mean from certain values. However, I don't want to do this over the whole column, just over a small range. This is my code at present:

data = pd.DataFrame({"test":[12, 4, 5, 4, 1, 3, 2, 5, 10, 9, 4, 3, 2, 1, 4, 2, 2, 4, 2, 5]})
range_start = 5
range_finish = 17

mask = np.arange(len(data)) %4
measured_stress_ratio_overload = data.iloc[range_start:range_finish, mask == 0, 'test'].mean()
measured_stress_ratio_baseline = data.iloc[range_start:range_finish, mask!= 0, 'test'].mean()

My expected output would be that I gain the average of the values at position 8, 12, 16 for measured stress_ratio_overload, and measured_stress_ratio_baseline all the other values between 5 and 17. However, when I try to run this code, I get this error:

IndexingError: Too many indexers

How do I use this range to properly index and retrieve the answer I'd like? Any help would be greatly appreciated!

pandas mask numpy
1 Answers

You shouldn't put the mask in the iloc. Since you are using divisor as a standard to find your desired row. You can first add a new column in your dataframe and then slice it.

data['divisor'] = np.arange(len(data)) %4

measured_stress_ratio_overload = data.iloc[range_start:range_finish][data['divisor'] == 0]['test'].mean()

measured_stress_ratio_baseline = data[data['divisor'] != 0].iloc[range_start:range_finish]['test'].mean()

or you can use df.where

measured_stress_ratio_overload = data.iloc[range_start:range_finish].where(data['divisor'] == 0)['test'].mean()

measured_stress_ratio_baseline = data.iloc[range_start:range_finish].where(data['divisor'] != 0)['test'].mean()
user_168360780
Related