I have a column in a pandas dataframe for genres. It is a string with the genres seperated by a column.
>>> df['genres_omdb']
0 Crime, Drama
1 Adventure, Family, Fantasy
2 Drama, Mystery
3 Horror, Mystery, Thriller
5 Action, Adventure, Sci-Fi
6 Drama, Romance
8 Drama
9 Animation, Adventure, Comedy
10 Animation, Adventure, Comedy
11 Drama, Sci-Fi
12 Drama
13 Drama, Romance, War
14 Comedy, Drama, Family
16 Comedy, Musical, Romance
So originally I split it into three columns and ran get_dummies on each of the columns. This produced repetitive columns (i.e. genre1_Adventure genre2_Adventure).
So then I tried getting every unique genre, creating a column of that genre, and then manually iterating through the rows and changing values to a 1 if the genre is in the list.
genre1_keys = df['genre1'].value_counts().keys()
genre2_keys = df['genre2'].value_counts().keys()
genre3_keys = df['genre3'].value_counts().keys()
for genre in genre1_keys:
all_genres.add(genre.strip())
for genre in genre2_keys:
all_genres.add(genre.strip())
for genre in genre3_keys:
all_genres.add(genre.strip())
for genre in all_genres:
df[genre] = 0
for i, row in df.iterrows():
genres = row['genres_omdb'].split(',')
for genre in genres:
genre = genre.strip()
row[genre] = 1
It's very messy and I know there is a better way to do this. Any help on how to clean up this code would be appreciated.
