问题
I have the following dataset:
import pandas as pd
import numpy as np
%matplotlib inline
df = pd.DataFrame({'movie' : ['A', 'B','C','D'],
'genres': ['Science Fiction|Romance|Family', 'Action|Romance',
'Family|Drama','Mystery|Science Fiction|Drama']},
index=range(4))
df
My attempt
# Parse unique genre from all the movies
gen = []
for g in df['genres']:
gg = g.split('|')
gen = gen + gg
gen = list(set(gen))
print(gen)
df['genres'].value_counts().plot(kind='pie')
I got this image:
But I would like to pie chart for each separate genres.
How we get the genres for number count of movies for each unique genres?
回答1:
You can do .str.split()
with expand=True
, which will give you a DataFrame
of all the genres. If you then stack that, you will get the value counts for all of the genres.
df.genres.str.split('|', expand=True).stack().value_counts().plot(kind='pie', label='Genre')
That can be a bit on the slower side to calculate the counts, so a faster implementation for the same plot would be (adding the percentages):
from itertools import chain
from collections import Counter
import matplotlib.pyplot as plt
cts = Counter(chain.from_iterable(df.genres.str.split('|').values))
_ = plt.pie(cts.values(), labels=cts.keys(), autopct='%1.0f%%')
_ = plt.ylabel('Genres')
回答2:
So, the one-liner solution:
df.genres.str.get_dummies().sum().plot.pie(label='Genre', autopct='%1.0f%%')
Result:
TL;DR
Firstly, convert your categories column to dummies:
df = pd.concat([df.drop('genres', axis=1), df.genres.str.get_dummies()], axis=1)
Result:
movie a b c d e f g
0 A 1 1 1 0 0 0 0
1 B 0 0 1 0 1 0 0
2 C 0 0 0 0 0 1 1
3 D 1 1 0 1 1 0 0
Then count number of occurrences for each category:
counts = df.drop('movie', axis=1).sum()
Result:
a 2
b 2
c 2
d 1
e 2
f 1
g 1
And finally plot the pie chart:
counts.plot.pie()
来源:https://stackoverflow.com/questions/52132970/pandas-how-to-plot-the-pie-diagram-for-the-movie-counts-versus-genre-of-imdb-mo