根据其他两个数据框列中的值条件创建新列

Question

I'm new to python. I have a feeling there is a quick fix, but nothing has seemed quick to me yet.

I have a 150,000+ row dataframe, data . Within it are two series: gridcode and CH4_Flux among others. I want to create a new categorical column, called category that assigns a categorical identifier to each row based on the four conditions shown here:

cat1 = data[(data.gridcode <= threshAV) & (data.CH4_Flux >= threshAM)]
cat2 = data[(data.gridcode >= threshAV) & (data.CH4_Flux >= threshAM)]
cat3 = data[(data.gridcode <= threshAV) & (data.CH4_Flux <= threshAM)]
cat4 = data[(data.gridcode >= threshAV) & (data.CH4_Flux <= threshAM)]

where threshAV is a predesignated threshold for gridcode , and threshAM is a predesignated threshold for CH4_Flux . Essentially either both exceed the threshold, neither exceed, or one or the other exceeds. Preferably the categorical labels would simply be the integers 1,2,3, and 4 respectively following the logic of cat1 , cat2 , cat3 ,and cat4 above.

I have tried for loops and if and where statements but have struck out.

When experimenting with for loops, I commonly get the error:

ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

Thank you in advance for any advice or direction!

Answer 1

Use numpy.select for new columns by multiple boolean masks:

#removed filtering by data[]
m1 = (data.gridcode <= threshAV) & (data.CH4_Flux >= threshAM)
m2 = (data.gridcode >= threshAV) & (data.CH4_Flux >= threshAM)
m3 = (data.gridcode <= threshAV) & (data.CH4_Flux <= threshAM)
m4 = (data.gridcode >= threshAV) & (data.CH4_Flux <= threshAM)

data['category'] = np.select([m1, m2, m3, m4], [1,2,3,4])

Or:

data['category'] = np.select([m1, m2, m3, m4], ['cat1','cat2','cat3','cat4'])