This is the pandas book version to deal with the overlapping categories

categorize_multiple2.py

'''Example data

    Name       Genre
0   TENET      Action|Thriller
1   MEMENTO    Crime|Thriller|Action
2   AVENGERS   Children's
'''
# SPOILER ALERT: This method is the UNDERLYING method. Just use - df.Genre.str.get_dummies("|") for the same result (more on this later)

# Step 1: Get the unique genre
gens = []
for gen in df.Genre:
    gens.extend(gen.split("|"))
gens = pd.unique(gens)


# Step 2: Construct the DF to store 0 and 1
zero_one = DataFrame(np.zeros(len(df.Name), len(gens), columns= gens))


# Step 3: MAIN - Use .get_indexer to get location for each movie's genre
for i, gen in enumerate(df.Genre):
    indices = zero_one.columns.get_indexer(gen.split("|"))
    zerp_one.iloc[i, indices] = 1
                     
# DONE!

'''Now MORE ON THIS LATER part:

The 
df.Genre.str.get_dummies("|") 
would have resulted the same but in the one line!

Do that... and if you want to learn the internals, then go for the written part!
'''
                 

You can see my own version of dealing with the overlapping categorical values here:
https://gist.github.com/AayushSameerShah/58e09fd89833f467dc462ba0807bf733