张凡幸，人工智能初阶小白，初学Python，喜欢数据分析、爬虫相关内容

微信公众号：学习随笔（ID：walker398）

# coding: utf-8
import pandas as pd
import requests
import random
import json
import time
header={
    'User-Agent':'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/55.0.2883.87 Safari/537.36',
    'Connection':'keep-alive'
}
cookies = '_lxsdk_cuid=1651950e6dfc8-02b4e01b2f82ec-6b1b1279-100200-1651950e6dfc8; _lxsdk=FA2EEC709AFE11E883FE1F23FDF\
          D6ADBB39CE732DF384D9DADFD1909B110AD64; v=3; _lx_utm=utm_source%3DBaidu%26utm_medium%3Dorganic; __mta=121983668.153372\
          8057235.1533731802598.1533731803845.10'
cookie = {}
for line in cookies.split(';'):
    name, value = cookies.strip().split('=', 1)
    cookie[name] = value
love_apartment = pd.DataFrame(columns=['data','score','city','comment','nick'])
for i in range(0,1000):
    j = random.randint(1,1000)
    print(str(i)+' '+str(j))
    try:
        time.sleep(2)
        url = 'https://m.maoyan.com/mmdb/comments/movie/1175253.json?_v_=yes&offset='+str(j)
        html = requests.get(url=url, cookies=cookie,headers=header).content
        data = json.loads(html.decode('utf-8'))['cmts']
        for item in data:
            love_apartment = love_apartment.append({'data':item['time'].split(' ')[0],'city':item['cityName'],
                                                    'score':str(item['score']),'comment':item['content'],'nick':item['nick']},
                                                   ignore_index=True)
        love_apartment.to_csv('爱情公寓.csv',index=False)
    except:
        continue

import pandas as pd
from pyecharts import Bar, Line,Overlap,Geo
import jieba
from wordcloud import WordCloud,ImageColorGenerator
import matplotlib.pyplot as plt
from collections import Counter

love_apartment_com = pd.read_excel('爱情公寓.xlsx',encoding='utf-8')
grouped = love_apartment_com.groupby(['city'])
grouped_pct = grouped['score']

city_com = grouped_pct.agg(['mean', 'count'])
city_com.reset_index(inplace=True)
city_com['mean'] = round(city_com['mean'], 2)
data = [(city_com['city'][i], city_com['count'][i]) for i in range(0, city_com.shape[0])]
geo = Geo('《爱情公寓》全国热力图', title_color="#fff",title_pos="center", width=1200, height=600, background_color='#404a59')
attr, value = geo.cast(data)
geo.add("", attr, value, type="heatmap",  is_visualmap=True, visual_range=[0, 20],visual_text_color="#fff")
geo.render('爱情公寓全国观影图.html')

在做第一张图表的时候，出现最大的问题就是pyecharts地图资源不全面，所以只好自己慢慢对Excel表筛除无效数据，着实是费了不少时间。从图中看出来，内陆确实人烟稀少，长三角，珠三角，沿海，北京，重庆都是金钱大佬，毕竟能去看电影，评得起分......

city_main = city_com.sort_values('count',ascending=False)[0:20]
attr = city_main['city']
v1 = city_main['count']
v2 = city_main['mean']
line = Line("主要城市评分")
line.add("城市", attr, v2, is_stack=True, xaxis_rotate=30, yaxis_min=0, mark_point=["max", "min"], xaxis_interval=0,
         line_color='lightblue', line_width=4, mark_point_textcolor='black', mark_point_color='lightblue', is_splitline_show=False)
bar = Bar("主要城市评论数")
bar.add("城市", attr, v1, is_stack=True, xaxis_rotate=30, yaxis_min=0, xaxis_interval=0, is_splitline_show=False)
overlap = Overlap()
overlap.add(bar)
overlap.add(line, yaxis_index=1, is_add_yaxis=True)
overlap.render('主要城市评分and评论数.html')

city_score = city_main.sort_values('mean',ascending=False)[0:20]
attr = city_score['city']
v1 = city_score['mean']
line = Line('爱情公寓主要城市评分')
line.add("城市", attr, v1, is_stack=True, xaxis_rotate=30, yaxis_min=0, mark_point=["min", "max"], xaxis_interval=0,
         line_color='lightblue', line_width=4, mark_point_textcolor='black', mark_point_color='lightblue', is_splitline_show=True)
line.render('爱情公寓主要城市评分.html')

city_score_area = city_com.sort_values('count',ascending=False)[0:30]city_score_area.reset_index(inplace=True)data = [(city_score_area['city'][i], city_score_area['mean'][i]) for i in range(0, city_score_area.shape[0])]geo = Geo('《爱情公寓》全国打分图', title_color="#fff",title_pos="center", width=1200, height=600, background_color='#404a59')attr, value = geo.cast(data)geo.add("", attr, value, visual_range=[0, 5],visual_text_color="#fff", symbol_size=15, is_visualmap=True,is_roam=False)geo.render('爱情公寓全国打分图.html')

love_apartment_str = ' '.join(love_apartment_com['comment'])
words_list = []
word_generator = jieba.cut_for_search(love_apartment_str)
for word in word_generator:
    words_list.append(word)
words_list = [k for k in words_list if len(k)>1]
back_color = plt.imread('爱心.jpg')
wc = WordCloud(background_color='white',
               mask=back_color,
               font_path='C:\Windows\Fonts\STZHONGS.TTF',
               max_words=2000,
               max_font_size=250,
               random_state=30,
               )
love_apartment_count = Counter(words_list)
wc.generate_from_frequencies(love_apartment_count)
image_colors = ImageColorGenerator(back_color)
plt.figure()
plt.imshow(wc.recolor(color_func=image_colors))
plt.axis('off')
print('生成词云成功!')
wc.to_file('love_apartment.jpg')