Chapter 3: ROC

[tanthiamhuat]

sorry for the many questions, because I could not get through the code. How do you get ROC about 0.879? What are the details inside the function area_under_curve? I suppose we need to have the data for: cv_data, b_all_artist_ids, model.transform to be input for the area_under_curve function. But even that, I had error for b_all_artist_ids

[srowen]

@analyticalmonk will have to supply the full code listings. That line looks like a bug in the listing unfortunately - it can't be broadcasting an int. I assume the .count() should be removed on the line before.

[tanthiamhuat]

but it looks to me that the function 'area_under_curve' needs to be defined too, else how can it be used?

[srowen]

Yes, that's what I said?

[analyticalmonk]

@tanthiamhuat the area_under_curve code is available in the chapter 3 notebook on the pyspark branch.