当前位置:

首页 > temp > python入门教程 >

Python分析5000+抖音大V，发现大家都喜欢这类视频！

写在前面

最近，我在知乎上看到一个关于抖音的问题。

里面提到了，目前我国人均每天刷短视频110分钟。

看这数据，看来我又被平均了。

不过老实说，只要一打开抖音，确实是有一种停不下来的感觉～

所以还是少刷抖音，多看书，多写代码。要不然时间全流逝了。

这次给大家用数据分析一下在抖音上，什么类型的视频最受欢迎。
Jupyter代码,数据可视化&分析

"""
当然在学习Python的道路上肯定会困难，没有好的学习资料，怎么去学习呢？
学习Python中有不明白推荐加入交流Q群号：928946953
群里有志同道合的小伙伴，互帮互助，群里有不错的视频学习教程和PDF！
还有大牛解答！
"""

from pyecharts.charts import Pie, Bar, TreeMap, Map, Geo
from wordcloud import WordCloud, ImageColorGenerator
from pyecharts import options as opts
from pyecharts.globals import ThemeType
import matplotlib.pyplot as plt
from PIL import Image
import pandas as pd
import numpy as np
import jieba

In [2]:

df = pd.read_csv('../file/douyin.csv',encoding = 'utf-8-sig')
df.head()

Out[2]:

name

gender

country

province

city

location

category

fans

videos

likes

comments

shares

following

school

custom_verify

enterprise_verify

signature

人民日报

中国

北京

政企

117259000

2427

1165446000

11906782048

9089061412

NaN

人民日报官方账号

参与、沟通、记录时代。

央视新闻

中国

北京

政企

105648000

3681

3814571666

2603872833

1989050522

NaN

央视新闻官方账号

央视新闻官方抖音号

本宝宝暂时还没想到个性签名

陈赫

中国

NaN

上海

明星

68374000

422

570096000

430908721

117639297

131

上海戏剧学院

演员陈赫

NaN

帅到没朋友 ‍♂️有东西直播间 1⃣️1⃣️1⃣️8⃣️ ☑️公众号➰[陈赫]

Dear-迪丽热巴

NaN

明星

49790000

181167000

202448645

151645265

NaN

演员

NaN

毒舌电影

中国

广东

广州

剧情

46355000

616

820393000

28026109

13005392

NaN

优质影视自媒体、抖音影评团成员

NaN

看电影，可以改变人生。商务邮：dsmovie@youhaoxi.cn ❤️ 日历预售...

In [27]:

df.loc[df.gender == '0', 'gender'] = '未知'
df.loc[df.gender == '1', 'gender'] = '男性'
df.loc[df.gender == '2', 'gender'] = '女性'
# 根据性别分组
gender_message = df.groupby(['gender'])
# 对分组后的结果进行计数
gender_com = gender_message['gender'].agg(['count'])
gender_com.reset_index(inplace=True)

# 饼图数据
attr = gender_com['gender']
v1 = gender_com['count']

# 初始化配置
pie = Pie(init_opts=opts.InitOpts(width="800px", height="400px",theme=ThemeType.LIGHT))
# 添加数据，设置半径
pie.add("", [list(z) for z in zip(attr, v1)], radius=["40%", "75%"])
# 设置全局配置项，标题、图例、工具箱(下载图片)
pie.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V性别分布情况", pos_left="center", pos_top="top"),
                    legend_opts=opts.LegendOpts(orient="vertical", pos_left="left"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}))
# 设置系列配置项，标签样式
pie.set_series_opts(label_opts=opts.LabelOpts(is_show=True, formatter="{b}:{d}%",font_size=14))
pie.render_notebook()

Out[27]:

In [26]:

df = df.sort_values('likes', ascending=False)
# 获取TOP10的数据
attr = df['name'][0:10]
v1 = [float('%.1f' % (float(i) / 100000000)) for i in df['likes'][0:10]]

# 初始化配置
bar = Bar(init_opts=opts.InitOpts(width="1000px", height="600px"))
# x轴数据
bar.add_xaxis(list(reversed(attr.tolist())))
# y轴数据
bar.add_yaxis("", list(reversed(v1)),color = '#84E0E3')
# 设置全局配置项，标题、y轴分割线
bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V点赞数TOP10(亿)", pos_left="center", pos_top="18"),
                    xaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)),
                    yaxis_opts=opts.AxisOpts(axislabel_opts=opts.LabelOpts(font_size=12))
                   )
# 设置系列配置项，标签样式
bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="right", color="black"))
bar.reversal_axis()
bar.render_notebook()

Out[26]:

In [29]:

				
										# 将数据分段

										Bins = [0, 1000000, 5000000, 10000000, 25000000, 50000000, 100000000, 5000000000]

										Labels = ['0-100', '100-500', '500-1000', '1000-2500', '2500-5000', '5000-10000', '10000以上']

										len_stage = pd.cut(df['likes'], bins=Bins, labels=Labels).value_counts().sort_index()

										# 获取数据

										attr = len_stage.index.tolist()

										v1 = len_stage.values.tolist()

										# 生成柱状图

										bar = Bar(init_opts=opts.InitOpts(width="800px", height="400px"))

										bar.add_xaxis(attr)

										bar.add_yaxis("", v1,color = '#84E0E3')

										bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V点赞数分布情况(万)", pos_left="center", pos_top="18"),

										                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),

										                    yaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)))

										bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="top", color="black"))

										bar.render_notebook()

Out[29]:

In [34]:

df = df.sort_values('fans', ascending=False)
attr = df['name'][0:10]
v1 = ['%.1f' % (float(i) / 10000) for i in df['fans'][0:10]]

bar = Bar(init_opts=opts.InitOpts(width="1000px", height="600px"))
bar.add_xaxis(list(reversed(attr.tolist())))
bar.add_yaxis("", list(reversed(v1)),color = '#84E0E3')
bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V粉丝数TOP10(万)", pos_left="center", pos_top="18"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                    xaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)))
bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="right", color="black"))
bar.reversal_axis()
bar.render_notebook()

Out[34]:

In [37]:

Bins = [0, 1500000, 2000000, 5000000, 10000000, 25000000, 200000000]
Labels = ['0-150', '150-200', '200-500', '500-1000', '1000-2500', '5000以上']
len_stage = pd.cut(df['fans'], bins=Bins, labels=Labels).value_counts().sort_index()

attr = len_stage.index.tolist()
v1 = len_stage.values.tolist()

bar = Bar(init_opts=opts.InitOpts(width="800px", height="400px"))
bar.add_xaxis(attr)
bar.add_yaxis("", v1,color = '#84E0E3')
bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V粉丝数分布情况(万)", pos_left="center", pos_top="18"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                    yaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)))
bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="top", color="black"))
bar.render_notebook()

Out[37]:

In [40]:

df = df.sort_values('comments', ascending=False)
attr = df['name'][0:10]
v1 = ['%.1f' % (float(i) / 100000000) for i in df['comments'][0:10]]

bar = Bar(init_opts=opts.InitOpts(width="1000px", height="600px"))
bar.add_xaxis(list(reversed(attr.tolist())))
bar.add_yaxis("", list(reversed(v1)),color = '#84E0E3')
bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V评论数TOP10(亿)", pos_left="center", pos_top="18"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                    xaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)))
bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="right", color="black"))
bar.reversal_axis()
bar.render_notebook()

Out[40]:

In [44]:

df = df.sort_values('shares', ascending=False)
attr = df['name'][0:10]
v1 = ['%.1f' % (float(i) / 100000000) for i in df['shares'][0:10]]

bar = Bar(init_opts=opts.InitOpts(width="1000px", height="600px"))
bar.add_xaxis(list(reversed(attr.tolist())))
bar.add_yaxis("", list(reversed(v1)),color = '#84E0E3')

bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V分享数TOP10(亿)", pos_left="center", pos_top="18"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                    xaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)),
                    yaxis_opts=opts.AxisOpts(axislabel_opts=opts.LabelOpts(rotate=30))
                   )
bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="right", color="black"))
bar.reversal_axis()
bar.render_notebook()

Out[44]:

In [28]:

# 分组求和
likes_type_message = df.groupby(['category'])
likes_type_com = likes_type_message['likes'].agg(['sum'])
likes_type_com.reset_index(inplace=True)
# 处理数据
dom = [{'name':name, 'value':num} for name, num in zip(likes_type_com['category'], likes_type_com['sum'])]

# 初始化配置
treemap = TreeMap(init_opts=opts.InitOpts(width="1000px", height="600px",theme=ThemeType.LIGHT))
# 添加数据
treemap.add('', dom)
# 设置全局配置项，标题、工具箱(下载图片)
treemap.set_global_opts(title_opts=opts.TitleOpts(title="各类型抖音大V点赞数汇总图", pos_left="center", pos_top="5"),
                        toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                        legend_opts=opts.LegendOpts(is_show=False),
                        
                       )
treemap.render_notebook()

Out[28]:

In [32]:

dom = []
fans_type_message = df.groupby(['category'])
fans_type_com = fans_type_message['fans'].agg(['sum'])
fans_type_com.reset_index(inplace=True)
for name, num in zip(fans_type_com['category'], fans_type_com['sum']):
    data = {}
    data['name'] = name
    data['value'] = num
    dom.append(data)

treemap = TreeMap(init_opts=opts.InitOpts(width="1000px", height="600px",theme=ThemeType.LIGHT))
treemap.add('', dom)
treemap.set_global_opts(title_opts=opts.TitleOpts(title="各类型抖音大V粉丝数汇总图", pos_left="center", pos_top="5"),
                        toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                        legend_opts=opts.LegendOpts(is_show=False))
treemap.set_series_opts(treemapbreadcrumb_opts=opts.TreeMapBreadcrumbOpts(is_show=False))
treemap.render_notebook()

Out[32]:

In [4]:

# 筛选
df = df[df['videos'] > 0]
# 计算单个视频平均点赞数
df.eval('result = likes/(videos*10000)', inplace=True)
df['result'] = df['result'].round(decimals=1)
df = df.sort_values('result', ascending=False)

# 取TOP10
attr = df['name'][0:10]
v1 = ['%.1f' % (float(i)) for i in  df['result'][0:10]]

# 初始化配置
bar = Bar(init_opts=opts.InitOpts(width="1000px", height="600px"))
# 添加数据
bar.add_xaxis(list(reversed(attr.tolist())))
bar.add_yaxis("", list(reversed(v1)),color = '#84E0E3')
# 设置全局配置项，标题、工具箱(下载图片)、y轴分割线
bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V平均视频点赞数TOP10(万)", pos_left="center", pos_top="18"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                    xaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)))
# 设置系列配置项
bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="right", color="black"))
# 翻转xy轴
bar.reversal_axis()
bar.render_notebook()

Out[4]:

In [13]:

# 筛选数据
df = df[df["country"] == "中国"]
df1 = df.copy()
# 数据替换
df1["province"] = df1["province"].str.replace("省", "").str.replace("壮族自治区", "").str.replace("维吾尔自治区", "").str.replace("自治区", "")
# 分组计数
df_num = df1.groupby("province")["province"].agg(count="count")
df_province = df_num.index.values.tolist()
df_count = df_num["count"].values.tolist()

# 初始化配置
map = Map(init_opts=opts.InitOpts(width="1000px", height="600px"))
# 中国地图
map.add("", [list(z) for z in zip(df_province, df_count)], "china")
# 设置全局配置项，标题、工具箱(下载图片)、颜色图例
map.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V省份分布情况", pos_left="center", pos_top="0"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                    # 设置数值范围0-600，is_piecewise标签值连续
                    visualmap_opts=opts.VisualMapOpts(max_=600, is_piecewise=False))
map.render_notebook()

Out[13]:

In [17]:

df1 = df[(df["school"] != "") & (df["school"] != "已毕业") & (df["school"] != "未知")]
df1 = df1.copy()
df_num = df1.groupby("school")["school"].agg(count="count").reset_index().sort_values(by="count", ascending=False)
df_school = df_num[:10]["school"].values.tolist()
df_count = df_num[:10]["count"].values.tolist()

# 初始化配置
bar = Bar(init_opts=opts.InitOpts(width="1200px", height="400px"))
bar.add_xaxis(df_school)
bar.add_yaxis("", df_count,color = '#84E0E3')
bar.set_global_opts(title_opts=opts.TitleOpts(title="抖音大V毕业学校TOP10", pos_left="center", pos_top="18"),
                    toolbox_opts=opts.ToolboxOpts(is_show=True, feature={"saveAsImage": {}}),
                    yaxis_opts=opts.AxisOpts(splitline_opts=opts.SplitLineOpts(is_show=True)))
bar.set_series_opts(label_opts=opts.LabelOpts(is_show=True, position="top", color="black"))
bar.render_notebook()

Out[17]:

In [21]:

"""
生成标题以及摘要词云
"""
words = pd.read_csv('../file/chineseStopWords.txt', encoding='gbk', sep='\t', names=['stopword'])
# 分词
text = ''
df1 = df[df["signature"] != ""]
df1 = df1.copy()
for line in df1['signature']:
    text += ' '.join(jieba.cut(str(line).replace(" ", ""), cut_all=False))
# 停用词
stopwords = set('')
stopwords.update(words['stopword'])
backgroud_Image = plt.imread('../file/douyin.png')
# 使用抖音背景色
alice_coloring = np.array(Image.open(r"../file/douyin.png"))
image_colors = ImageColorGenerator(alice_coloring)
wc = WordCloud(
    background_color='white',
    mask=backgroud_Image,
    font_path='../file/simhei.ttf',
    max_words=2000,
    max_font_size=70,
    min_font_size=1,
    prefer_horizontal=1,
    color_func=image_colors,
    random_state=50,
    stopwords=stopwords,
    margin=5
)
wc.generate_from_text(text)
wc.to_file('../file/douyin_word.png')
print('生成词云成功!')
点击并拖拽以移动
Building prefix dict from the default dictionary ...
Loading model from cache C:\Users\10076\AppData\Local\Temp\jieba.cache
Loading model cost 0.550 seconds.
Prefix dict has been built successfully.
c:\users\10076\appdata\local\programs\python\python38\lib\site-packages\wordcloud\wordcloud.py:995: UserWarning: mask image should be unsigned byte between 0 and 255. Got a float array
  warnings.warn("mask image should be unsigned byte between 0"

生成词云成功!

数据获取

数据来自于第三方监测，一共是有5000+抖音大V的数据信息(所需的文件我会放在文章末,供大家下载)

主要包含大V们的昵称、性别、地点、类型、点赞数、粉丝数、视频数、评论数、分享数、关注数、毕业学校、认证、简介等信息。

其中粉丝最多的是「人民日报」，接近1.2亿。「央视新闻」也破亿了，记得之前破亿的时候还上过热搜～

粉丝最少的博主也有近150w+的粉丝，这5000多位大V累计236.5亿粉丝，地球人口的三倍多！
(由于时间的原因,这些大V的粉丝数肯定比这还要高)

数据可视化&分析

导入第三方库,然后读取数据

from pyecharts.charts import Pie, Bar, TreeMap, Map, Geo
from wordcloud import WordCloud, ImageColorGenerator
from pyecharts import options as opts
from pyecharts.globals import ThemeType
import matplotlib.pyplot as plt
from PIL import Image
import pandas as pd
import numpy as np
import jieba

df = pd.read_csv('../file/douyin.csv',encoding = 'utf-8-sig')
df.head()