import numpy as np
import pandas as pd#数据处理
import jieba
import wordcloud
import collections
#图形显示
import matplotlib.pyplot as plt
import seaborn as sns
from pyecharts.charts import Map
from pyecharts import options as opts
from pyecharts.globals import ThemeType,SymbolType,ChartType
from PIL import Image
# 让表格显示全年面
pd.set_option('display.max_columns',None)
pd.set_option('display.max_rows',None)
pd.set_option('display.expand_frame_repr',False)
pd.set_option('display.float_format',lambda x:'%.2f' % x)
# 画板设置
sns.set()
colors=sns.color_palette()
# print(colors)
# 中文配置
plt.rcParams['font.sans-serif']='SimHei'
plt.rcParams['axes.unicode_minus']=False
zxy=Image.open(r'D:\pyproject\excel\84期录播课配套资料\day08-电影舆情分析\代码\zxy.jfif')
df=pd.read_excel(r'D:\pyproject\excel\白蛇(含省份)数据.xlsx')
# print(df.isnull().mean())
df.dropna(how='any',axis=0)
df.index=np.arange(df.shape[0])
# 评分处理
df_scores=df['评分'].value_counts().sort_index(ascending=False)
x=np.arange(df_scores.size)
# print(x)
df_scores.index.map(lambda x:f'{x}分')
figure=plt.figure(figsize=(15,5))
plt.bar(x,df_scores.values)
plt.xticks(x,df_scores.index.map(lambda x:f'{x}分'))
son_figure=figure.add_subplot(1,3,3)
son_figure.pie(df_scores.values,labels=df_scores.index,autopct='%.1f%%')
plt.savefig('./5.jpg')
#评论量分析
df=df.set_index('评论时间')
comment_count=df.resample('D')['评论'].count()
# print(comment_count)
plt.figure(figsize=(12,5),alpha=1)
plt.plot(comment_count.index.day.tolist(),comment_count.values,color='green',marker='o')
plt.title('8月每日评论量',fontsize=12,color='green')
plt.fill_between(comment_count.index.day.tolist(),comment_count.values,color='green',alpha=0.6)
for x,y in zip(comment_count.index.day.tolist(),comment_count.values):
plt.text(x,y*1.05,str(y))
# print(comment_count.index.day.tolist)
plt.xticks(comment_count.index.day.tolist())
plt.savefig('./4.jpg')
#按小时评论量
df.reset_index(inplace=True)
df['小时']=df['评论时间'].dt.hour
time_count=df.groupby('小时')['评论'].count()
plt.figure(figsize=(12,5))
plt.title('每小时评论量',fontsize=12,color='red')
plt.bar(time_count.index,time_count.values)
plt.xticks(time_count.index,time_count.index.map(lambda x:f'{x}时'))
for x,y in zip(time_count.index,time_count.values):
plt.text(x,y,str(y))
plt.xlabel('时间')
plt.ylabel('评论次数')
plt.savefig('./3.jpg')
# 每周评论数
df['星期']=df['评论时间'].dt.dayofweek+1
week_count=df.groupby('星期')['评论'].count()
dic={
1:'星期一',
2:'星期二',
3:'星期三',
4:'星期四',
5:'星期五',
6:'星期六',
7:'星期日'
}
plt.figure(figsize=(12,5))
plt.plot(week_count.index,week_count.values,color=colors[3])
plt.bar(week_count.index,week_count.values,color=colors[5])
plt.xticks(week_count.index,week_count.index.map(dic))
plt.ylim(0,5000)
plt.title('一周每天评论量分析')
plt.savefig('./2.jpg')
# # 人物热度排序
# print(df.head())
name=['小青','小白','蒙面人','司马','牛头帮主','宝青坊主','法海','蜘蛛精']
comments=''.join(df['评论'].values)
list1=[]
for item in name:
num=comments.count(item)
list1.append((item,num))
person_count=pd.DataFrame(data=list1,columns=('名字','数量'))
person_count.sort_values('数量',inplace=True,ascending=False)
# print(person_count)
plt.figure(figsize=(12,5))
plt.bar(person_count['名字'],person_count['数量'])
# print(person_count)
plt.savefig('./1.jpg')
# 地图显示数据
data=df['省份'].value_counts()
c=(
Map(init_opts=opts.InitOpts())
.add("",[list(z) for z in zip(data.index.tolist(),data.values.tolist())],"china")
.set_global_opts(
title_opts=opts.TitleOpts(title="评论数量的地域分布"),
visualmap_opts=opts.VisualMapOpts(max_=2209,min_=53,range_color=['white','green'])
)
)
c.render_notebook()
c.render("./test.html")
#词云分析
comment_word=[]
stop_words=[]
with open(r'D:\pyproject\excel\84期录播课配套资料\day08-电影舆情分析\代码\stopwords.txt','r',encoding='utf-8') as f:
stop_word_1=f.readlines()
for stop_word in stop_word_1:
stop_word.strip('\n')
stop_words.append(stop_word)
for item in df['评论']:
word=jieba.lcut(item)
for good_word in word:
if good_word not in stop_words:
comment_word.append(good_word)
c=collections.Counter(comment_word)
plt.figure(figsize=(12,5))
wc=wordcloud.WordCloud(font_path=r'D:\pyproject\excel\84期录播课配套资料\day08-电影舆情分析\代码\蒙纳超刚黑简.ttf',
width=500,height=300,
background_color='white',
max_font_size=200,
min_font_size=5,
max_words=2000,
mask=np.array(zxy))
image=wc.generate_from_frequencies(c) #加入文字渲染
plt.axis('off')
plt.imshow(image)
plt.show()
评论10