一个小例子，对一份电影数据文件作简单分析，并统计电影类别的出现次数
- 读入这份电影数据集：
import numpy as np
import pandas as pd
from matplotlib import pyplot as plt
df = pd.read_csv("IMDB-Movie-Data.csv")
- 取电影类别数据，并对Genre类别字段的值以,为分隔构建数组：
genre_data = df["Genre"].str.split(",")
print(genre_data.head(5))
打印：
0 [Action, Adventure, Sci-Fi]
1 [Adventure, Mystery, Sci-Fi]
2 [Horror, Thriller]
3 [Animation, Comedy, Family]
4 [Action, Adventure, Fantasy]
Name: Genre, dtype: object
- 取所有电影类别：
genre_list = list()
for genre in genre_data:
genre_list.extend(genre)
genre_series = pd.Series(genre_list).unique()
print(genre_series)
打印：
['Action' 'Adventure' 'Sci-Fi' 'Mystery' 'Horror' 'Thriller' 'Animation'
'Comedy' 'Family' 'Fantasy' 'Drama' 'Music' 'Biography' 'Romance'
'History' 'Crime' 'Western' 'War' 'Musical' 'Sport']
- 以电影为行，以类别为列，构建以0填充的DataFrame：
matrix = np.zeros((len(genre_data), len(genre_series)), int)
matrix = pd.DataFrame(matrix, columns=genre_series)
print(matrix.head(5))
打印：
0 0 0 0 0 ... 0 0 0 0
1 0 0 0 0 ... 0 0 0 0
2 0 0 0 0 ... 0 0 0 0
3 0 0 0 0 ... 0 0 0 0
4 0 0 0 0 ... 0 0 0 0
[5 rows x 20 columns]
- 给矩阵赋值
for index in range(len(genre_data)):
genre = genre_data[index]
# loc方法可以传入列名数组，选出所有列
matrix.loc[index, genre] = 1
print(matrix.head(5))
打印：
Action Adventure Sci-Fi Mystery ... Western War Musical Sport
0 1 1 1 0 ... 0 0 0 0
1 0 1 1 1 ... 0 0 0 0
2 0 0 0 0 ... 0 0 0 0
3 0 0 0 0 ... 0 0 0 0
4 1 1 0 0 ... 0 0 0 0
[5 rows x 20 columns]
- 统计电影各类别出现次数
genre_count = matrix.sum(axis=0)
print(genre_count)
打印：
Action 303
Adventure 259
Sci-Fi 120
Mystery 106
Horror 119
Thriller 195
Animation 49
Comedy 279
Family 51
Fantasy 101
Drama 513
Music 16
Biography 81
Romance 141
History 29
Crime 150
Western 7
War 13
Musical 5
Sport 18
dtype: int64
- 绘制条形图
plt.figure(figsize=(20, 8), dpi=80)
_x = genre_count.index
_y = genre_count.values
plt.bar(_x, _y)
plt.show()
电影类别频数统计
源码
Github: MoviesDataAnalyze
