TMDB电影数据分析

1.简介

数据基本信息

本数据集中包含 1 万余条电影信息，信息来源为“电影数据库”（TMDb，The Movie Database），包括21个特征指标：id、热度、票房、预算、片名、演职人员、导演、类型、用户评分、评分人数、发行时间等21个特征。*
时间：1960-2015年

主要分析数据

热度（popularity）、预算（budget），电影类型（genre），上映时间（release_date），票房收入（revenue ），平均评分（vote_average），评分次数（vote_count）*

分析内容

分析工具：python、Power BI

2.分析结果

2.1 电影类型（市场分布）

2.2 票房

2.3 档期

2.4 观众评价（评分&评价人数）

2.5 观众喜好（热度）

2.6 盈利情况

3. 结论

1、电影类型：

• 总体来看，电影数量从2000年开始快速增长，其中剧情片、喜剧片、惊悚片、动作片、恐怖片、科幻片、纪录片增长最快。
• 2005-2015年，喜剧、奇幻、爱情、家庭片的比例呈下降趋势，音乐类、恐怖片、纪录片、科幻片、惊悚片的比例上升。
• 在20种电影类型中，数量排名前六的电影分别是：剧情、喜剧、惊悚、动作、爱情、恐怖。数量最少的是西部片、非院线电影、外国电影、历史片、战争片。

2、票房：

• 票房总体呈上升趋势，1986年后票房增长加快，之后变化较平稳，2006后快速增长。
• 票房排名前六的电影类型分别是：动画、奇幻、冒险、家庭、科幻、动作，其变化趋势较平缓。其中冒险、动作呈下降趋势。 而纪录片、恐怖片、外国片、剧情片、历史片的票房较低，应该谨慎拍摄。
• 剧情片的票房增长最快，其次是惊悚片、恐怖片、喜剧、犯罪片。
• 电影票房与受欢迎度，评价次数，电影预算均成正相关。增加预算可更好地保证电影质量和后期宣传，有助于获得更多票房。

3、档期：

• 票房排名前五的月份：6月、5月、11月、7月。票房较低的月份是9月、1月、8月。
  9月、10月、12月、1月发行的影片最多，竞争比较激烈。
• 根据各月份的票房和竞争度，5月、6月、7月、11月是电影发行的最佳档期。8月、9月、1月须谨慎发行。12月虽然票房较高，但竞争激烈，挑战与机遇并存。
• 周五上映的影片最多，星期一、星期二与星期三上映的票房最高，远高于休息日。

4、评分：

• 评分人数整体呈上升趋势，从2007年开始增加，2013年后略有下降。
• 70年代开始，随着电影产出量的增加，影片层次不齐，低分影片开始出现，平均分也有略微下降，人们对影片的要求越来越高。
• 各类型电影的平均评分在5-7分，差距不大。可以看出，电影类型对评分影响不是很大。同时，类型、评价数量、票房、热度等对电影评分的影响都不大，电影评分的高低主要还是取决于观众的主观评价、电影的质量等。

5、观众喜好（热度）

• 热度排名前六的类型分别为：科幻、冒险、奇幻、动画、动作、家庭。想象题材或非现实题材的电影（科幻、冒险、奇幻、动画）的受欢迎度要明显高于其他类别的电影。
• 电影热度与评分数、票房、预算成正相关，但也有预算小、热度高的电影。增加预算可更好地保证电影质量和后期宣传，有助于增加热度。

6、电影盈利情况：

• 利润：排名前六的类型：动画、奇幻、冒险、家庭、科幻、动作。纪录片、西部片、恐怖片、历史片的利润最低。
• 投资回报率：排名前六的类型：纪录片、音乐类电影、爱情片、家庭片、恐怖片。西部片、历史片的投资回报率较低，亏损风险较高，应该谨慎拍摄。

7、分析局限性：

• 60%的数据存在票房、预算缺失，去除异常值后数据集仅有3000多条，分析准确度可能有所下降。

4.数据处理过程（python）

# 导入模块
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

%matplotlib inline
plt.style.use('ggplot')    

3.1 数据清洗

常规属性

# 加载数据并打印几行。进行这几项操作，来检查数据
df = pd.read_csv('tmdb_movies.csv')
df.head()

	id	imdb_id	popularity	budget	revenue	original_title	cast	homepage	director	tagline	...	overview	runtime	genres	production_companies	release_date	vote_count	vote_average	release_year	budget_adj	revenue_adj
0	135397	tt0369610	32.985763	150000000	1513528810	Jurassic World	Chris Pratt|Bryce Dallas Howard|Irrfan Khan|Vi...	http://www.jurassicworld.com/	Colin Trevorrow	The park is open.	...	Twenty-two years after the events of Jurassic ...	124	Action|Adventure|Science Fiction|Thriller	Universal Studios|Amblin Entertainment|Legenda...	6/9/15	5562	6.5	2015	1.379999e+08	1.392446e+09
1	76341	tt1392190	28.419936	150000000	378436354	Mad Max: Fury Road	Tom Hardy|Charlize Theron|Hugh Keays-Byrne|Nic...	http://www.madmaxmovie.com/	George Miller	What a Lovely Day.	...	An apocalyptic story set in the furthest reach...	120	Action|Adventure|Science Fiction|Thriller	Village Roadshow Pictures|Kennedy Miller Produ...	5/13/15	6185	7.1	2015	1.379999e+08	3.481613e+08
2	262500	tt2908446	13.112507	110000000	295238201	Insurgent	Shailene Woodley|Theo James|Kate Winslet|Ansel...	http://www.thedivergentseries.movie/#insurgent	Robert Schwentke	One Choice Can Destroy You	...	Beatrice Prior must confront her inner demons ...	119	Adventure|Science Fiction|Thriller	Summit Entertainment|Mandeville Films|Red Wago...	3/18/15	2480	6.3	2015	1.012000e+08	2.716190e+08
3	140607	tt2488496	11.173104	200000000	2068178225	Star Wars: The Force Awakens	Harrison Ford|Mark Hamill|Carrie Fisher|Adam D...	http://www.starwars.com/films/star-wars-episod...	J.J. Abrams	Every generation has a story.	...	Thirty years after defeating the Galactic Empi...	136	Action|Adventure|Science Fiction|Fantasy	Lucasfilm|Truenorth Productions|Bad Robot	12/15/15	5292	7.5	2015	1.839999e+08	1.902723e+09
4	168259	tt2820852	9.335014	190000000	1506249360	Furious 7	Vin Diesel|Paul Walker|Jason Statham|Michelle ...	http://www.furious7.com/	James Wan	Vengeance Hits Home	...	Deckard Shaw seeks revenge against Dominic Tor...	137	Action|Crime|Thriller	Universal Pictures|Original Film|Media Rights ...	4/1/15	2947	7.3	2015	1.747999e+08	1.385749e+09

5 rows × 21 columns

#类型，以及是否有缺失数据或错误数据的情况
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10866 entries, 0 to 10865
Data columns (total 21 columns):
 #   Column                Non-Null Count  Dtype  
---  ------                --------------  -----  
 0   id                    10866 non-null  int64  
 1   imdb_id               10856 non-null  object 
 2   popularity            10866 non-null  float64
 3   budget                10866 non-null  int64  
 4   revenue               10866 non-null  int64  
 5   original_title        10866 non-null  object 
 6   cast                  10790 non-null  object 
 7   homepage              2936 non-null   object 
 8   director              10822 non-null  object 
 9   tagline               8042 non-null   object 
 10  keywords              9373 non-null   object 
 11  overview              10862 non-null  object 
 12  runtime               10866 non-null  int64  
 13  genres                10843 non-null  object 
 14  production_companies  9836 non-null   object 
 15  release_date          10866 non-null  object 
 16  vote_count            10866 non-null  int64  
 17  vote_average          10866 non-null  float64
 18  release_year          10866 non-null  int64  
 19  budget_adj            10866 non-null  float64
 20  revenue_adj           10866 non-null  float64
dtypes: float64(4), int64(6), object(11)
memory usage: 1.7+ MB

数据清理（清除多余列、丢空、去重。）

#清除多余列
df.drop(['imdb_id','cast','homepage', 'director', 'tagline', 'keywords', 'overview','production_companies'], axis = 1, inplace = True)
df.head(2)

	id	popularity	budget	revenue	original_title	runtime	genres	release_date	vote_count	vote_average	release_year	budget_adj	revenue_adj
0	135397	32.985763	150000000	1513528810	Jurassic World	124	Action|Adventure|Science Fiction|Thriller	6/9/15	5562	6.5	2015	1.379999e+08	1.392446e+09
1	76341	28.419936	150000000	378436354	Mad Max: Fury Road	120	Action|Adventure|Science Fiction|Thriller	5/13/15	6185	7.1	2015	1.379999e+08	3.481613e+08

#查看缺失值数量
df.isnull().sum()

id                 0
popularity         0
budget             0
revenue            0
original_title     0
runtime            0
genres            23
release_date       0
vote_count         0
vote_average       0
release_year       0
budget_adj         0
revenue_adj        0
dtype: int64

#去除含有任何空值的行
df.dropna(inplace = True)

#检查任何列是否还有空值
df.isnull().sum().any()

False

#查看数据重复数量
sum(df.duplicated())

1

#去除重复行
df.drop_duplicates(inplace = True)

#确认重复数据是否删除
sum(df.duplicated())

0

df.describe()

	id	popularity	budget	revenue	runtime	vote_count	vote_average	release_year	budget_adj	revenue_adj
count	10842.000000	10842.000000	1.084200e+04	1.084200e+04	10842.000000	10842.000000	10842.000000	10842.000000	1.084200e+04	1.084200e+04
mean	65870.675521	0.647461	1.465531e+07	3.991138e+07	102.138443	217.823649	5.974064	2001.314794	1.758712e+07	5.147797e+07
std	91981.355752	1.001032	3.093971e+07	1.171179e+08	31.294612	576.180993	0.934257	12.813617	3.433437e+07	1.447723e+08
min	5.000000	0.000065	0.000000e+00	0.000000e+00	0.000000	10.000000	1.500000	1960.000000	0.000000e+00	0.000000e+00
25%	10589.250000	0.208210	0.000000e+00	0.000000e+00	90.000000	17.000000	5.400000	1995.000000	0.000000e+00	0.000000e+00
50%	20557.000000	0.384532	0.000000e+00	0.000000e+00	99.000000	38.000000	6.000000	2006.000000	0.000000e+00	0.000000e+00
75%	75186.000000	0.715393	1.500000e+07	2.414118e+07	111.000000	146.000000	6.600000	2011.000000	2.092507e+07	3.387838e+07
max	417859.000000	32.985763	4.250000e+08	2.781506e+09	900.000000	9767.000000	9.200000	2015.000000	4.250000e+08	2.827124e+09

3.2 探索性数据分析

电影类型分析：各类型电影数量分布及随时间变化趋势如何？
（1）获取电影类型

# 创建一个集合,有去重功能
genres_set = set()
#切割genres列，创建一个循环
for i in df['genres']:
    genres_set.update(i.split('|'))
genres_set.discard(' ')
genres_set

{'Action',
 'Adventure',
 'Animation',
 'Comedy',
 'Crime',
 'Documentary',
 'Drama',
 'Family',
 'Fantasy',
 'Foreign',
 'History',
 'Horror',
 'Music',
 'Mystery',
 'Romance',
 'Science Fiction',
 'TV Movie',
 'Thriller',
 'War',
 'Western'}

#新建一个数据框
genres_df = pd.DataFrame()
#判断每部电影的类型，电影包含某个类型就返回1，否则返回0
for gen in genres_set:
    genres_df[gen] = df['genres'].str.contains(gen).map(lambda x: 1 if x else 0)
genres_df.head(10)    

	Music	Action	Thriller	Horror	War	Fantasy	TV Movie	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Foreign	Adventure	Western	History	Documentary	Romance
0	0	1	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0
1	0	1	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0
2	0	0	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0
3	0	1	0	0	0	1	0	0	1	0	0	0	0	0	0	1	0	0	0	0
4	0	1	1	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0
5	0	0	1	0	0	0	0	0	0	0	0	0	0	1	0	1	1	0	0	0
6	0	1	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0
7	0	0	0	0	0	0	0	0	1	0	0	0	0	1	0	1	0	0	0	0
8	0	0	0	0	0	0	0	1	0	1	0	0	1	0	0	1	0	0	0	0
9	0	0	0	0	0	0	0	1	0	1	0	0	1	0	0	0	0	0	0	0

#查看各类型电影数量
g1 = pd.DataFrame(index = [0,1])
for gen in genres_set:
    g1[gen] =genres_df[gen].value_counts()
    
g1

	Music	Action	Thriller	Horror	War	Fantasy	TV Movie	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Foreign	Adventure	Western	History	Documentary	Romance
0	10434	8458	7935	9205	10572	9926	10675	9611	9613	7049	10032	9488	10143	6082	10654	9371	10677	10508	10322	9130
1	408	2384	2907	1637	270	916	167	1231	1229	3793	810	1354	699	4760	188	1471	165	334	520	1712

#在数据框中加入年份
genres_df['release_year'] = df['release_year']
genres_df['release_year'] = df['release_year']
#数据框按年份分组，求和
gen_year = genres_df.groupby('release_year').sum()
gen_year.head()

	Music	Action	Thriller	Horror	War	Fantasy	TV Movie	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Foreign	Adventure	Western	History	Documentary	Romance
release_year
1960	1	8	6	7	2	2	0	3	3	8	0	2	0	13	1	5	6	5	0	6
1961	2	7	0	3	2	2	0	5	4	10	1	2	1	16	1	6	3	3	0	7
1962	1	8	7	5	3	1	0	2	2	5	4	3	0	21	1	7	3	4	0	5
1963	0	4	10	9	1	2	0	3	2	13	6	4	1	13	2	7	2	4	0	8
1964	5	5	9	6	3	4	0	4	4	16	4	10	2	20	1	5	1	3	0	9

（2）1960-2015年各类型电影数量变化

plt.figure(figsize = (12, 6))
plt.plot(gen_year, label = gen_year.columns)
plt.title('Number of movies by genres and year')
plt.xticks(range(1950, 2020, 5))
plt.xlabel('Year')
plt.ylabel('Moies Number')
plt.legend(gen_year)#图例

#每年各类型电影总数
genres_sum_year = gen_year.sum( axis=1)

#每年各类型电影占比
gen_proportion_year = pd.DataFrame()
for i in gen_year.columns:
    gen_proportion_year[i] = gen_year[i]/genres_sum_year
gen_proportion_year.head()

	Music	Action	Thriller	Horror	War	Fantasy	TV Movie	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Foreign	Adventure	Western	History	Documentary	Romance
release_year
1960	0.012821	0.102564	0.076923	0.089744	0.025641	0.025641	0.0	0.038462	0.038462	0.102564	0.000000	0.025641	0.000000	0.166667	0.012821	0.064103	0.076923	0.064103	0.0	0.076923
1961	0.026667	0.093333	0.000000	0.040000	0.026667	0.026667	0.0	0.066667	0.053333	0.133333	0.013333	0.026667	0.013333	0.213333	0.013333	0.080000	0.040000	0.040000	0.0	0.093333
1962	0.012195	0.097561	0.085366	0.060976	0.036585	0.012195	0.0	0.024390	0.024390	0.060976	0.048780	0.036585	0.000000	0.256098	0.012195	0.085366	0.036585	0.048780	0.0	0.060976
1963	0.000000	0.043956	0.109890	0.098901	0.010989	0.021978	0.0	0.032967	0.021978	0.142857	0.065934	0.043956	0.010989	0.142857	0.021978	0.076923	0.021978	0.043956	0.0	0.087912
1964	0.045045	0.045045	0.081081	0.054054	0.027027	0.036036	0.0	0.036036	0.036036	0.144144	0.036036	0.090090	0.018018	0.180180	0.009009	0.045045	0.009009	0.027027	0.0	0.081081

#2005-2015年各类型电影占比
gen_proportion_year1 = gen_proportion_year.loc[2005:,:]
gen_proportion_year1

	Music	Action	Thriller	Horror	War	Fantasy	TV Movie	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Foreign	Adventure	Western	History	Documentary	Romance
release_year
2005	0.013904	0.074866	0.098396	0.059893	0.005348	0.037433	0.003209	0.060963	0.026738	0.152941	0.031016	0.048128	0.027807	0.194652	0.010695	0.056684	0.002139	0.010695	0.018182	0.066310
2006	0.011494	0.076628	0.109195	0.053640	0.006705	0.032567	0.007663	0.062261	0.028736	0.148467	0.028736	0.050766	0.037356	0.188697	0.009579	0.052682	0.000958	0.011494	0.015326	0.067050
2007	0.013321	0.084369	0.111012	0.070160	0.005329	0.041741	0.005329	0.039964	0.036412	0.134103	0.030195	0.057726	0.028419	0.174956	0.015098	0.053286	0.004440	0.011545	0.016874	0.065719
2008	0.016142	0.079903	0.102502	0.061340	0.014528	0.034705	0.004036	0.045198	0.041969	0.136400	0.023406	0.050040	0.026634	0.188055	0.014528	0.050847	0.001614	0.019370	0.020985	0.067797
2009	0.012518	0.079529	0.115611	0.066274	0.008837	0.036082	0.005891	0.044183	0.052283	0.145803	0.037555	0.038292	0.035346	0.164948	0.012518	0.053019	0.000000	0.011782	0.018409	0.061119
2010	0.008258	0.088357	0.110652	0.064410	0.005780	0.036334	0.006606	0.045417	0.037159	0.139554	0.026424	0.041288	0.041288	0.173410	0.010735	0.048720	0.004955	0.011561	0.028902	0.070190
2011	0.013302	0.089984	0.114241	0.061033	0.007042	0.035994	0.007825	0.056338	0.043818	0.134585	0.029734	0.037559	0.035994	0.167449	0.010955	0.048513	0.002347	0.006260	0.038341	0.058685
2012	0.017255	0.077647	0.125490	0.081569	0.007843	0.026667	0.010980	0.032941	0.042353	0.138039	0.025882	0.042353	0.031373	0.181961	0.004706	0.039216	0.003137	0.010196	0.038431	0.061961
2013	0.021692	0.087491	0.126537	0.073753	0.005061	0.028200	0.007231	0.034707	0.044107	0.126537	0.027477	0.051338	0.030369	0.182936	0.000000	0.048445	0.002169	0.010123	0.044830	0.046999
2014	0.019048	0.087755	0.121769	0.071429	0.015646	0.024490	0.009524	0.029252	0.042177	0.125850	0.024490	0.044218	0.024490	0.193197	0.000000	0.045578	0.004082	0.010204	0.049660	0.057143
2015	0.023810	0.077201	0.123377	0.090188	0.006494	0.023810	0.014430	0.031746	0.062049	0.116883	0.030303	0.036797	0.028139	0.187590	0.000000	0.049784	0.004329	0.010823	0.041126	0.041126

（3）2005-2015年各类型电影占比变化

#2005-2015年各类型电影占比变化
plt.figure(figsize = (12, 6))
plt.plot(gen_proportion_year.loc[2005:, :], label = gen_year.columns)
plt.title('Proportion of movies by genres and year(2005-2015)')
plt.xticks(range(2003, 2016, 1))
plt.xlabel('Year')
plt.ylabel('Moies Number')
plt.legend(gen_proportion_year)#图例

（4）各类型电影比例分布

#各电影类型数量
genres_sum = genres_df.sum().sort_values(ascending = False).drop('release_year')
genres_sum

Drama              4760
Comedy             3793
Thriller           2907
Action             2384
Romance            1712
Horror             1637
Adventure          1471
Crime              1354
Family             1231
Science Fiction    1229
Fantasy             916
Mystery             810
Animation           699
Documentary         520
Music               408
History             334
War                 270
Foreign             188
TV Movie            167
Western             165
dtype: int64

genres_total = genres_sum.sum()
#各电影类型比例
genres_proportion = genres_sum/ genres_total
genres_proportion

Drama              0.176591
Comedy             0.140716
Thriller           0.107846
Action             0.088444
Romance            0.063513
Horror             0.060731
Adventure          0.054572
Crime              0.050232
Family             0.045669
Science Fiction    0.045595
Fantasy            0.033983
Mystery            0.030050
Animation          0.025932
Documentary        0.019291
Music              0.015136
History            0.012391
War                0.010017
Foreign            0.006975
TV Movie           0.006196
Western            0.006121
dtype: float64

#绘制柱形图
genres_proportion.plot.barh(label = 'genre', figsize = (12, 6))
plt.title('Proportion of Gennres')
plt.xlabel('Proportion')
plt.ylabel('Genre')

进一步做数据清洗，将清洗后的数据放在df2数据集中：

筛选评分人数大于50的数据。评分人数（vote_count）过低的电影，其评分（vote_average）不具有统计意义。
筛去票房、预算等为0的数据。

#筛选评分人数大于50的数据
df2 = df.query('vote_count > 50')
#筛去票房、预算等为0的数据
df2=df2[~df2['budget'].isin([0.000000e+00])]
df2=df2[~df2['revenue'].isin([0.000000e+00])]
df2.head()

	id	popularity	budget	revenue	original_title	runtime	genres	release_date	vote_count	vote_average	release_year	budget_adj	revenue_adj
0	135397	32.985763	150000000	1513528810	Jurassic World	124	Action|Adventure|Science Fiction|Thriller	6/9/15	5562	6.5	2015	1.379999e+08	1.392446e+09
1	76341	28.419936	150000000	378436354	Mad Max: Fury Road	120	Action|Adventure|Science Fiction|Thriller	5/13/15	6185	7.1	2015	1.379999e+08	3.481613e+08
2	262500	13.112507	110000000	295238201	Insurgent	119	Adventure|Science Fiction|Thriller	3/18/15	2480	6.3	2015	1.012000e+08	2.716190e+08
3	140607	11.173104	200000000	2068178225	Star Wars: The Force Awakens	136	Action|Adventure|Science Fiction|Fantasy	12/15/15	5292	7.5	2015	1.839999e+08	1.902723e+09
4	168259	9.335014	190000000	1506249360	Furious 7	137	Action|Crime|Thriller	4/1/15	2947	7.3	2015	1.747999e+08	1.385749e+09

df2.describe()

	id	popularity	budget	revenue	runtime	vote_count	vote_average	release_year	budget_adj	revenue_adj
count	3123.000000	3123.000000	3.123000e+03	3.123000e+03	3123.000000	3123.000000	3123.000000	3123.000000	3.123000e+03	3.123000e+03
mean	42688.892731	1.392294	4.261919e+07	1.296105e+08	110.037464	644.723343	6.254371	2002.292027	4.968112e+07	1.629720e+08
std	71354.906196	1.569972	4.461333e+07	1.891287e+08	19.588370	939.891310	0.760518	10.935365	4.697721e+07	2.311534e+08
min	5.000000	0.010335	1.000000e+00	2.000000e+00	26.000000	51.000000	3.300000	1960.000000	9.693980e-01	2.861934e+00
25%	4000.000000	0.617888	1.300000e+07	2.496112e+07	96.000000	135.000000	5.700000	1997.000000	1.657964e+07	3.145014e+07
50%	10585.000000	0.976612	2.800000e+07	6.556987e+07	106.000000	299.000000	6.300000	2005.000000	3.463336e+07	8.285793e+07
75%	44927.500000	1.593279	5.950000e+07	1.556332e+08	120.000000	717.500000	6.800000	2011.000000	6.911341e+07	1.967146e+08
max	336004.000000	32.985763	4.250000e+08	2.781506e+09	248.000000	9767.000000	8.400000	2015.000000	4.250000e+08	2.827124e+09

df2.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 3123 entries, 0 to 10822
Data columns (total 13 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   id              3123 non-null   int64  
 1   popularity      3123 non-null   float64
 2   budget          3123 non-null   int64  
 3   revenue         3123 non-null   int64  
 4   original_title  3123 non-null   object 
 5   runtime         3123 non-null   int64  
 6   genres          3123 non-null   object 
 7   release_date    3123 non-null   object 
 8   vote_count      3123 non-null   int64  
 9   vote_average    3123 non-null   float64
 10  release_year    3123 non-null   int64  
 11  budget_adj      3123 non-null   float64
 12  revenue_adj     3123 non-null   float64
dtypes: float64(4), int64(6), object(3)
memory usage: 341.6+ KB

2、票房分析：票房和哪些特征有关？

（1）电影票房的影响因素（数值型变量）

#绘制相关系数热力图
plt.subplots(figsize=(8,8))#调节图像大小
sns.heatmap(df2.corr(), annot = True, vmax = 1, square = True, cmap = 'Reds' )

可以看出，电影票房和评价次数（0.74）、预算（0.67）、受欢迎度（0.59）相关性较强。

绘制电影票房与受欢迎度，评价次数，电影预算的相关性散点图及其线性回归线

#创建票房与受欢迎度，评价次数，电影预算的数据框
revenue = df2[['vote_count', 'budget', 'popularity', 'revenue']]
plt.figure(figsize = (18,6))

#电影票房与受欢迎度的相关性散点图及其线性回归线
ax1 = plt.subplot(1, 3, 1)
ax1 = sns.regplot(x = 'popularity', y = 'revenue', data = revenue, color = 'y')
ax1.text(0, 2.5, 'r = 0.59')
plt.title('Revenue and Popularity')
plt.xlabel('Popularity')
plt.ylabel('Revenue')

#电影票房与评价次数的相关性散点图及其线性回归线
ax2 = plt.subplot(1, 3, 2)
ax2 = sns.regplot(x ='vote_count' , y = 'revenue', data = revenue, color = 'r')
ax2.text(0, 2.5, 'r = 0.74')
plt.title('Revnnue and Vote count')
plt.xlabel('Vote count')
plt.ylabel('Revenue')

#电影票房与预算的相关性散点图及其线性回归线
ax3 = plt.subplot(1, 3, 3)
ax3 = sns.regplot(x = 'budget' , y ='revenue', data = revenue, color = 'B')
ax3.text(0, 2.5, 'r = 0.67')
plt.title('Revnnue and Budget ')
plt.xlabel('Budget')
plt.ylabel('Revenue')

（2）票房和电影类型

#新建一个数据框
genres_df2 = pd.DataFrame()
#判断每部电影的类型，电影包含某个类型就返回1，否则返回0
for gen in genres_set:
    genres_df2[gen] = df2['genres'].str.contains(gen).map(lambda x: 1 if x else 0)
genres_df2.head()    

	Music	Action	Thriller	Horror	War	Fantasy	TV Movie	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Foreign	Adventure	Western	History	Documentary	Romance
0	0	1	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0
1	0	1	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0
2	0	0	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0
3	0	1	0	0	0	1	0	0	1	0	0	0	0	0	0	1	0	0	0	0
4	0	1	1	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0

#查看各类型电影数量
g0 = pd.DataFrame(index = [0,1])
for gen in genres_set:
    g0[gen] =genres_df2[gen].value_counts()
    
g0

	Music	Action	Thriller	Horror	War	Fantasy	TV Movie	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Foreign	Adventure	Western	History	Documentary	Romance
0	3023	2198	2095	2747	3023	2769	3123.0	2758	2671	2058	2836	2570	2937	1768	3122	2446	3084	3014	3108	2625
1	100	925	1028	376	100	354	NaN	365	452	1065	287	553	186	1355	1	677	39	109	15	498

#删掉样本数量少的类型

genres_df2.drop(['TV Movie','Foreign'],axis = 1, inplace = True)
genres_df2.head()

	Music	Action	Thriller	Horror	War	Fantasy	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Adventure	Western	History	Documentary	Romance
0	0	1	1	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0
1	0	1	1	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0
2	0	0	1	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0
3	0	1	0	0	0	1	0	1	0	0	0	0	0	1	0	0	0	0
4	0	1	1	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0

#创建电影类型与票房、评分、预算、热度、评价数量数据框
genres_df3 = pd.DataFrame()
genres_df3 = pd.concat([genres_df2, df2.iloc[:, [1, 2, 3, 7, 8, 9, 10]]], axis = 1)
genres_df3.head()

	Music	Action	Thriller	Horror	War	Fantasy	Family	Science Fiction	Comedy	Mystery	...	History	Documentary	Romance	popularity	budget	revenue	release_date	vote_count	vote_average	release_year
0	0	1	1	0	0	0	0	1	0	0	...	0	0	0	32.985763	150000000	1513528810	6/9/15	5562	6.5	2015
1	0	1	1	0	0	0	0	1	0	0	...	0	0	0	28.419936	150000000	378436354	5/13/15	6185	7.1	2015
2	0	0	1	0	0	0	0	1	0	0	...	0	0	0	13.112507	110000000	295238201	3/18/15	2480	6.3	2015
3	0	1	0	0	0	1	0	1	0	0	...	0	0	0	11.173104	200000000	2068178225	12/15/15	5292	7.5	2015
4	0	1	1	0	0	0	0	0	0	0	...	0	0	0	9.335014	190000000	1506249360	4/1/15	2947	7.3	2015

5 rows × 25 columns

genres_set.remove('TV Movie')
genres_set.remove('Foreign')
genres_set

{'Action',
 'Adventure',
 'Animation',
 'Comedy',
 'Crime',
 'Documentary',
 'Drama',
 'Family',
 'Fantasy',
 'History',
 'Horror',
 'Music',
 'Mystery',
 'Romance',
 'Science Fiction',
 'Thriller',
 'War',
 'Western'}

分别计算不同电影的平均评分、平均受欢迎度、平均票房

#创建三个数组，
vote_by_genre = pd.Series(index = genres_set)
pop_by_genre = pd.Series(index = genres_set)
rev_by_genre = pd.Series(index = genres_set)
bud_by_genre = pd.Series(index = genres_set)

#分别计算不同电影的平均评分、平均受欢迎度、平均票房
for gen in genres_set:
    vote_by_genre[gen] = genres_df3.groupby(gen, as_index = False).mean().loc[1, 'vote_average']
    pop_by_genre[gen] = genres_df3.groupby(gen, as_index = False).mean().loc[1, 'popularity']
    rev_by_genre[gen] = genres_df3.groupby(gen, as_index = False).mean().loc[1, 'revenue']
    bud_by_genre[gen] = genres_df3.groupby(gen, as_index = False).mean().loc[1, 'budget']
#合并三个数组
movie_by_genre = pd.DataFrame({ 'vote_average': vote_by_genre, 'popularity': pop_by_genre, 'revenue': rev_by_genre, 'budget': bud_by_genre})

movie_by_genre

	vote_average	popularity	revenue	budget
Music	6.385000	1.101270	1.047237e+08	3.013560e+07
Action	6.099568	1.777369	1.802467e+08	6.309883e+07
Thriller	6.143093	1.413200	1.123881e+08	4.111380e+07
Horror	5.834840	0.974864	6.870872e+07	2.116128e+07
War	6.691000	1.412921	1.238682e+08	4.698285e+07
Fantasy	6.135876	1.916209	2.413179e+08	7.649616e+07
Family	6.210137	1.638098	2.320525e+08	6.898803e+07
Science Fiction	6.107522	2.097370	1.865328e+08	6.238452e+07
Comedy	6.115305	1.200185	1.200283e+08	3.844007e+07
Mystery	6.283275	1.300413	1.071573e+08	3.850171e+07
Crime	6.366004	1.261440	9.649637e+07	3.647491e+07
Animation	6.454839	1.821527	2.740895e+08	8.015984e+07
Drama	6.519188	1.197703	9.213394e+07	3.224194e+07
Adventure	6.177400	2.027486	2.392040e+08	7.520216e+07
Western	6.541026	1.407328	1.119227e+08	6.475754e+07
History	6.659633	1.085134	9.390964e+07	4.473635e+07
Documentary	6.726667	0.449032	4.494842e+07	8.142680e+06
Romance	6.317470	1.163230	1.097685e+08	3.171751e+07

不同类型电影的平均票房图

movie_by_genre.sort_values(by = ['revenue'])['revenue'].plot.barh(figsize = (10, 5))
plt.title('Average revenue by Genres')
plt.xlabel('Revenue')
plt.ylabel('Genres')

（3）电影票房随时间变化趋势

plt.figure(figsize = (10, 5))
revenue_by_year = df2.groupby('release_year').mean().sort_values(by = 'release_year')['revenue']
plt.plot(revenue_by_year)

plt.title('Average revenue by Release year')
plt.xlabel('Release year')
plt.ylabel('Average revenue')

（4）各类型票房的变化趋势

#每年各类型电影的票房
revnue_genre_year = pd.DataFrame(index = gen_proportion_year.index)

for gen in genres_set:
    revnue_genre_year[gen] = genres_df3.groupby(['release_year', gen]).revenue.mean().xs(0, level = 1)
    
revnue_genre_year.head()    

	Music	Action	Thriller	Horror	War	Fantasy	Family	Science Fiction	Comedy	Mystery	Crime	Animation	Drama	Adventure	Western	History	Documentary	Romance
release_year
1960	30476250.0	2.850000e+07	2.996833e+07	2.996833e+07	3.047625e+07	30476250.0	30476250.00	3.047625e+07	3.230167e+07	30476250.0	3.047625e+07	3.047625e+07	4.905000e+06	3.900000e+07	3.900000e+07	2.063500e+07	3.047625e+07	3.230167e+07
1961	66070003.5	6.975921e+07	6.158737e+07	6.158737e+07	6.975921e+07	61587367.2	23014205.50	6.158737e+07	2.751894e+07	61587367.2	6.607000e+07	2.301421e+07	2.158800e+08	2.105227e+07	6.158737e+07	7.448421e+07	6.158737e+07	7.460921e+07
1962	37682461.5	3.037662e+07	3.037662e+07	3.768246e+07	2.690995e+07	37682461.5	37682461.50	3.768246e+07	3.768246e+07	37682461.5	4.586667e+07	3.768246e+07	3.380000e+07	1.056492e+07	4.757662e+07	2.690995e+07	3.768246e+07	3.768246e+07
1963	33305376.4	2.190703e+07	3.457676e+07	3.878084e+07	4.038172e+07	33305376.4	33305376.40	3.330538e+07	3.826307e+07	38263073.5	3.330538e+07	3.330538e+07	3.459229e+07	2.754271e+07	3.330538e+07	3.459229e+07	3.330538e+07	3.176743e+07
1964	47113424.0	3.834271e+07	4.617839e+07	5.276892e+07	6.143466e+07	42868164.2	35585205.25	5.276892e+07	6.796667e+07	61922709.6	6.192271e+07	5.276892e+07	7.605776e+07	3.834271e+07	5.276892e+07	5.276892e+07	5.276892e+07	4.892271e+07

#用0填充nan值
revnue_genre_year = revnue_genre_year.fillna(0)
  

plt.figure(figsize = (12, 6))
plt.plot(revnue_genre_year, label = revnue_genre_year.columns)
plt.title('Average revenue by genres and year')
plt.xticks(range(1950, 2020, 5))
plt.xlabel('Year')
plt.ylabel('Average revenue')
plt.legend(revnue_genre_year)#图例

3.档期

（1）不同月份的发行数量和票房

genres_df3['release_date'] = pd.to_datetime(genres_df3['release_date'])


genres_df3['month'] = genres_df3['release_date'].dt.month
genres_df3['day'] = genres_df3['release_date'].dt.weekday
genres_df3.head()

	Music	Action	Thriller	Horror	War	Fantasy	Family	Science Fiction	Comedy	Mystery	...	Romance	popularity	budget	revenue	release_date	vote_count	vote_average	release_year	month	day
0	0	1	1	0	0	0	0	1	0	0	...	0	32.985763	150000000	1513528810	2015-06-09	5562	6.5	2015	6	1
1	0	1	1	0	0	0	0	1	0	0	...	0	28.419936	150000000	378436354	2015-05-13	6185	7.1	2015	5	2
2	0	0	1	0	0	0	0	1	0	0	...	0	13.112507	110000000	295238201	2015-03-18	2480	6.3	2015	3	2
3	0	1	0	0	0	1	0	1	0	0	...	0	11.173104	200000000	2068178225	2015-12-15	5292	7.5	2015	12	1
4	0	1	1	0	0	0	0	0	0	0	...	0	9.335014	190000000	1506249360	2015-04-01	2947	7.3	2015	4	2

5 rows × 27 columns

plt.figure(figsize = (18,6))

#各月份的电影票房
revenue_by_month = genres_df3.groupby('month')['revenue'].mean()
plt.subplot(1, 2, 1)
revenue_by_month.plot(kind = 'bar')
plt.title('Average revenue by Month')
plt.xlabel('Month')
plt.ylabel('Average revenue')

#各月份的电影发行数
df['release_date'] = pd.to_datetime(df['release_date'])


df['month'] = df['release_date'].dt.month
num_by_month = df.groupby('month')['revenue'].count()
plt.subplot(1, 2, 2)
num_by_month.plot(kind = 'bar')
plt.title('Number of movie by Month')
plt.xlabel('Month')
plt.ylabel('Number of movie')

（2）不同星期的发行数量和票房

plt.figure(figsize = (18,6))

#各月份的电影票房
revenue_by_month = genres_df3.groupby('day')['revenue'].mean()
plt.subplot(1, 2, 1)
revenue_by_month.plot(kind = 'bar')
plt.title('Average revenue by Day of week')
plt.xlabel('Day of week')
plt.ylabel('Average revenue')

#各月份的电影发行数
df['release_date'] = pd.to_datetime(df['release_date'])

df['day'] = df['release_date'].dt.weekday#返回0—6，分别对应星期一到星期日
num_by_month = df.groupby('day')['revenue'].count()
plt.subplot(1, 2, 2)
num_by_month.plot(kind = 'bar')
plt.title('Number of movie by Day of week')
plt.xlabel('Day od week')
plt.ylabel('Number of movie')

4、观众评价： 电影的评分与哪些特征有关？

(1)评分与电影类型

movie_by_genre.sort_values(by = ['vote_average'])['vote_average'].plot.barh(figsize = (10, 5))
plt.title('Average vote by Genres')
plt.xlabel('Average vote')
plt.ylabel('Genres')

（2）评分与其他变量

plt.subplots(figsize=(8,8))#调节图像大小
sns.heatmap(df.corr(), annot = True, vmax = 1, square = True, cmap = 'Reds' )

（3）评分人数的变化趋势（2000-2015）

#不同年代的评分变化
plt.figure(figsize = (10,6))
vote_count_by_year = genres_df3.groupby('release_year')['vote_count'].sum().loc[2000:]
vote_count_by_year.plot(kind = 'bar')
plt.title('Vote count by Year(2000-2015)')
plt.xlabel('Year')
plt.ylabel('Vote count')

（4）不同年代的评分变化

#绘制箱线图
s1960 = genres_df3.query('release_year <= 1970').vote_average
s1970 = genres_df3.query('release_year > 1970').query('release_year <= 1980').vote_average
s1980 = genres_df3.query('release_year > 1980').query('release_year <= 1990').vote_average
s1990 = genres_df3.query('release_year >1990').query('release_year <= 2000').vote_average
s2000 = genres_df3.query('release_year > 2000').query('release_year <= 2010').vote_average
s2010 = genres_df3.query('release_year > 2010').query('release_year <= 2020').vote_average

plt.figure(figsize = (12, 6))
plt.boxplot([s1960,s1970,s1980,s1990,s2000,s2010],labels = ['1960s', '1970s', '1980s', '1990s', '2000s', '2010s'])
plt.title('Average Vote by Decade')
plt.xlabel('Decade')
plt.ylabel('Average Vote')

4、热度分析：电影热度与哪些特征有关？

（1）热度与电影类型

movie_by_genre.sort_values(by = ['popularity'])['popularity'].plot.barh(figsize = (10, 5))
plt.title('Popularity by Genres')
plt.xlabel('Popularity')
plt.ylabel('Genres')
plt.show()

（2）受欢迎度与其他变量

plt.subplots(figsize=(8,8))#调节图像大小
sns.heatmap(df2.corr(), annot = True, vmax = 1, square = True, cmap = 'Blues' )

绘制电影热度与评价次数、票房、预算的相关性散点图及其线性回归线

#创建电影受欢迎度与评价次数、票房、预算的数据框
poularity = df2[['vote_count', 'budget', 'popularity', 'revenue']]
plt.figure(figsize = (18,6))

#电影受欢迎度与评价次数的相关性散点图及其线性回归线
ax1 = plt.subplot(1, 3, 1)
ax1 = sns.regplot(x = 'vote_count', y = 'popularity', data = poularity, color = 'y')
ax1.text(0, 25, 'r = 0.76')
plt.title('Popularity and Vote_count')
plt.xlabel('Vote_count')
plt.ylabel('Popularity')

#电影受欢迎度与票房的相关性散点图及其线性回归线
ax2 = plt.subplot(1, 3, 2)
ax2 = sns.regplot(x ='revenue' , y = 'popularity', data = poularity, color = 'r')
ax2.text(0, 25, 'r = 0.59')
plt.title('Popularity and Revenue')
plt.xlabel('Revenue')
plt.ylabel('Popularity')

#电影受欢迎度与预算的相关性散点图及其线性回归线
ax3 = plt.subplot(1, 3, 3)
ax3 = sns.regplot(x = 'budget' , y ='popularity', data = poularity, color = 'B')
ax3.text(0, 25, 'r = 0.52')
plt.title('Popularity and Budget ')
plt.xlabel('Budget')
plt.ylabel('Popularity')

5、电影盈利情况分析：电影利润与哪些因素有关？

(1)电影利润与类型

#计算各类型电影利润
movie_by_genre['profit'] = movie_by_genre['revenue'] - movie_by_genre['budget']
movie_by_genre

	vote_average	popularity	revenue	budget	profit
Music	6.385000	1.101270	1.047237e+08	3.013560e+07	7.458808e+07
Action	6.099568	1.777369	1.802467e+08	6.309883e+07	1.171479e+08
Thriller	6.143093	1.413200	1.123881e+08	4.111380e+07	7.127430e+07
Horror	5.834840	0.974864	6.870872e+07	2.116128e+07	4.754744e+07
War	6.691000	1.412921	1.238682e+08	4.698285e+07	7.688537e+07
Fantasy	6.135876	1.916209	2.413179e+08	7.649616e+07	1.648217e+08
Family	6.210137	1.638098	2.320525e+08	6.898803e+07	1.630644e+08
Science Fiction	6.107522	2.097370	1.865328e+08	6.238452e+07	1.241483e+08
Comedy	6.115305	1.200185	1.200283e+08	3.844007e+07	8.158827e+07
Mystery	6.283275	1.300413	1.071573e+08	3.850171e+07	6.865555e+07
Crime	6.366004	1.261440	9.649637e+07	3.647491e+07	6.002145e+07
Animation	6.454839	1.821527	2.740895e+08	8.015984e+07	1.939297e+08
Drama	6.519188	1.197703	9.213394e+07	3.224194e+07	5.989200e+07
Adventure	6.177400	2.027486	2.392040e+08	7.520216e+07	1.640018e+08
Western	6.541026	1.407328	1.119227e+08	6.475754e+07	4.716515e+07
History	6.659633	1.085134	9.390964e+07	4.473635e+07	4.917329e+07
Documentary	6.726667	0.449032	4.494842e+07	8.142680e+06	3.680574e+07
Romance	6.317470	1.163230	1.097685e+08	3.171751e+07	7.805097e+07

movie_by_genre.profit.sort_values().plot.barh(figsize = (10, 5))

plt.title('Profit by Genres')
plt.xlabel('Profit')
plt.ylabel('Genres')

（2）各类型的投资回报率（ROI）

movie_by_genre['ROI'] = movie_by_genre['profit']/movie_by_genre['budget']
movie_by_genre

	vote_average	popularity	revenue	budget	profit	ROI
Music	6.385000	1.101270	1.047237e+08	3.013560e+07	7.458808e+07	2.475082
Action	6.099568	1.777369	1.802467e+08	6.309883e+07	1.171479e+08	1.856577
Thriller	6.143093	1.413200	1.123881e+08	4.111380e+07	7.127430e+07	1.733586
Horror	5.834840	0.974864	6.870872e+07	2.116128e+07	4.754744e+07	2.246908
War	6.691000	1.412921	1.238682e+08	4.698285e+07	7.688537e+07	1.636456
Fantasy	6.135876	1.916209	2.413179e+08	7.649616e+07	1.648217e+08	2.154640
Family	6.210137	1.638098	2.320525e+08	6.898803e+07	1.630644e+08	2.363663
Science Fiction	6.107522	2.097370	1.865328e+08	6.238452e+07	1.241483e+08	1.990050
Comedy	6.115305	1.200185	1.200283e+08	3.844007e+07	8.158827e+07	2.122480
Mystery	6.283275	1.300413	1.071573e+08	3.850171e+07	6.865555e+07	1.783182
Crime	6.366004	1.261440	9.649637e+07	3.647491e+07	6.002145e+07	1.645554
Animation	6.454839	1.821527	2.740895e+08	8.015984e+07	1.939297e+08	2.419287
Drama	6.519188	1.197703	9.213394e+07	3.224194e+07	5.989200e+07	1.857581
Adventure	6.177400	2.027486	2.392040e+08	7.520216e+07	1.640018e+08	2.180812
Western	6.541026	1.407328	1.119227e+08	6.475754e+07	4.716515e+07	0.728335
History	6.659633	1.085134	9.390964e+07	4.473635e+07	4.917329e+07	1.099180
Documentary	6.726667	0.449032	4.494842e+07	8.142680e+06	3.680574e+07	4.520101
Romance	6.317470	1.163230	1.097685e+08	3.171751e+07	7.805097e+07	2.460817

#各类型电影的投资回报率
movie_by_genre['ROI'].sort_values().plot.barh(figsize = (10, 5))

plt.title('ROI by Genres')
plt.xlabel('Profit')
plt.ylabel('Genres')

(3) 电影利润与其他变量

df2['profit'] = df2['revenue'] - df2['budget']
plt.subplots(figsize=(8,8))#调节图像大小
sns.heatmap(df2.corr(), annot = True, vmax = 1, square = True, cmap = 'Reds' )

绘制电影利润与评价数、受欢迎度的散点图和趋势线

#创建利润与评价次数，受欢迎度的数据框
profit = df2[['vote_count', 'popularity', 'profit']]
plt.figure(figsize = (18,6))

#电影利润与评价次数的相关性散点图及其线性回归线
ax1 = plt.subplot(1, 2, 1)
ax1 = sns.regplot(x = 'vote_count', y = 'profit', data = profit, color = 'y')
ax1.text(0, 2.5, 'r = 0.71')
plt.title('Profit and Vote_count')
plt.xlabel('Vote_count')
plt.ylabel('Profit')

#电影利润与受欢迎度的相关性散点图及其线性回归线
ax2 = plt.subplot(1, 2, 2)
ax2 = sns.regplot(x ='popularity' , y = 'profit', data = profit, color = 'r')
ax2.text(0, 2.5, 'r = 0.57')
plt.title('Profit and Popularity')
plt.xlabel('Popularity')
plt.ylabel('Profit')