用Python分析了75w条数据，看看美国人更想让谁当总统？

https://tianchi.aliyun.com/competition/entrance/531837/introduction

# 安装词云处理包wordcloud
pip install wordcloud
# 数据可视化包matplotlib
pip install matplotlib
# 数据处理包pandas
pip install pandas

https://tianchi.aliyun.com/dataset/dataDetail?dataId=79412

https://tianchi.aliyun.com/competition/entrance/531837/introduction

# 导入相关处理包
import pandas as pd

# 读取候选人信息，由于原始数据没有表头，需要添加表头
candidates = pd.read_csv("weball20.txt", sep = '|',names=['CAND_ID','CAND_NAME','CAND_ICI','PTY_CD','CAND_PTY_AFFILIATION','TTL_RECEIPTS',
                                                          'TRANS_FROM_AUTH','TTL_DISB','TRANS_TO_AUTH','COH_BOP','COH_COP','CAND_CONTRIB',
                                                          'CAND_LOANS','OTHER_LOANS','CAND_LOAN_REPAY','OTHER_LOAN_REPAY','DEBTS_OWED_BY',
                                                          'TTL_INDIV_CONTRIB','CAND_OFFICE_ST','CAND_OFFICE_DISTRICT','SPEC_ELECTION','PRIM_ELECTION','RUN_ELECTION'
                                                          ,'GEN_ELECTION','GEN_ELECTION_PRECENT','OTHER_POL_CMTE_CONTRIB','POL_PTY_CONTRIB',
                                                          'CVG_END_DT','INDIV_REFUNDS','CMTE_REFUNDS'])
# 读取候选人和委员会的联系信息
ccl = pd.read_csv("ccl.txt", sep = '|',names=['CAND_ID','CAND_ELECTION_YR','FEC_ELECTION_YR','CMTE_ID','CMTE_TP','CMTE_DSGN','LINKAGE_ID'])

# 关联两个表数据
ccl = pd.merge(ccl,candidates)
# 提取出所需要的列
ccl = pd.DataFrame(ccl, columns=[ 'CMTE_ID','CAND_ID', 'CAND_NAME','CAND_PTY_AFFILIATION'])

# 查看目前ccl数据前10行
ccl.head(10)

# 读取个人捐赠数据，由于原始数据没有表头，需要添加表头
# 提示：读取本文件大概需要5-10s
itcont = pd.read_csv('itcont_2020_20200722_20200820.txt', sep='|',names=['CMTE_ID','AMNDT_IND','RPT_TP','TRANSACTION_PGI',
                                                                                  'IMAGE_NUM','TRANSACTION_TP','ENTITY_TP','NAME','CITY',
                                                                                  'STATE','ZIP_CODE','EMPLOYER','OCCUPATION','TRANSACTION_DT',
                                                                                  'TRANSACTION_AMT','OTHER_ID','TRAN_ID','FILE_NUM','MEMO_CD',
                                                                                  'MEMO_TEXT','SUB_ID'])

/opt/conda/lib/python3.6/site-packages/IPython/core/interactiveshell.py:3058: DtypeWarning: Columns (10,15,16,18) have mixed types. Specify dtype option on import or set low_memory=False.
  interactivity=interactivity, compiler=compiler, result=result)

# 将候选人与委员会关系表ccl和个人捐赠数据表itcont合并，通过 CMTE_ID
c_itcont =  pd.merge(ccl,itcont)
# 提取需要的数据列
c_itcont = pd.DataFrame(c_itcont, columns=[ 'CAND_NAME','NAME', 'STATE','EMPLOYER','OCCUPATION',
                                           'TRANSACTION_AMT', 'TRANSACTION_DT','CAND_PTY_AFFILIATION'])

# 查看目前数据前10行
c_itcont.head(10)

# 查看数据规模 多少行 多少列
c_itcont.shape

(756205, 8)

# 查看整体数据信息，包括每个字段的名称、非空数量、字段的数据类型
c_itcont.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 756205 entries, 0 to 756204
Data columns (total 8 columns):
CAND_NAME               756205 non-null object
NAME                    756205 non-null object
STATE                   756160 non-null object
EMPLOYER                737413 non-null object
OCCUPATION              741294 non-null object
TRANSACTION_AMT         756205 non-null int64
TRANSACTION_DT          756205 non-null int64
CAND_PTY_AFFILIATION    756205 non-null object
dtypes: int64(2), object(6)
memory usage: 51.9+ MB

#空值处理，统一填充 NOT PROVIDED
c_itcont['STATE'].fillna('NOT PROVIDED',inplace=True)
c_itcont['EMPLOYER'].fillna('NOT PROVIDED',inplace=True)
c_itcont['OCCUPATION'].fillna('NOT PROVIDED',inplace=True)

# 对日期TRANSACTION_DT列进行处理
c_itcont['TRANSACTION_DT'] = c_itcont['TRANSACTION_DT'] .astype(str)
# 将日期格式改为年月日  7242020 
c_itcont['TRANSACTION_DT'] = [i[3:7]+i[0]+i[1:3] for i in c_itcont['TRANSACTION_DT'] ]

# 再次查看数据信息
c_itcont.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 756205 entries, 0 to 756204
Data columns (total 8 columns):
CAND_NAME               756205 non-null object
NAME                    756205 non-null object
STATE                   756205 non-null object
EMPLOYER                756205 non-null object
OCCUPATION              756205 non-null object
TRANSACTION_AMT         756205 non-null int64
TRANSACTION_DT          756205 non-null object
CAND_PTY_AFFILIATION    756205 non-null object
dtypes: int64(1), object(7)
memory usage: 51.9+ MB

# 查看数据前3行
c_itcont.head(3)

# 查看数据表中数据类型的列的数据分布情况
c_itcont.describe()

# 查看单列的数据发布情况
c_itcont['CAND_NAME'].describe()

count                 756205
unique                   312
top       BIDEN, JOSEPH R JR
freq                  507816
Name: CAND_NAME, dtype: object

# 计算每个党派的所获得的捐款总额，然后排序，取前十位
c_itcont.groupby("CAND_PTY_AFFILIATION").sum().sort_values("TRANSACTION_AMT",ascending=False).head(10)

# 计算每个总统候选人所获得的捐款总额，然后排序，取前十位
c_itcont.groupby("CAND_NAME").sum().sort_values("TRANSACTION_AMT",ascending=False).head(10)

# 查看不同职业的人捐款的总额，然后排序，取前十位
c_itcont.groupby('OCCUPATION').sum().sort_values("TRANSACTION_AMT",ascending=False).head(10)

# 查看每个职业捐款人的数量
c_itcont['OCCUPATION'].value_counts().head(10)

NOT EMPLOYED    224109
RETIRED         151834
ATTORNEY         19666
NOT PROVIDED     14912
PHYSICIAN        14033
CONSULTANT        8333
PROFESSOR         8022
TEACHER           8013
ENGINEER          7922
SALES             6435
Name: OCCUPATION, dtype: int64

# 每个州获捐款的总额，然后排序，取前五位
c_itcont.groupby('STATE').sum().sort_values("TRANSACTION_AMT",ascending=False).head(5)

# 查看每个州捐款人的数量
c_itcont['STATE'].value_counts().head(5)

CA    127895
TX     54457
FL     54343
NY     49453
MA     29314
Name: STATE, dtype: int64

# 导入matplotlib中的pyplot
import matplotlib.pyplot as plt
# 为了使matplotlib图形能够内联显示
%matplotlib inline
# 导入词云库
from wordcloud import WordCloud,ImageColorGenerator

# 各州总捐款数可视化
st_amt = c_itcont.groupby('STATE').sum().sort_values("TRANSACTION_AMT",ascending=False)[:10]
st_amt=pd.DataFrame(st_amt, columns=['TRANSACTION_AMT'])
st_amt.plot(kind='bar')

<AxesSubplot:xlabel='STATE'>

# 各州捐款总人数可视化，取前10个州的数据
st_amt = c_itcont.groupby('STATE').size().sort_values(ascending=False).head(10)
st_amt.plot(kind='bar')

<AxesSubplot:xlabel='STATE'>

# 从所有数据中取出支持拜登的数据
biden = c_itcont[c_itcont['CAND_NAME']=='BIDEN, JOSEPH R JR']
# 统计各州对拜登的捐款总数
biden_state = biden.groupby('STATE').sum().sort_values("TRANSACTION_AMT", ascending=False).head(10)
# 饼图可视化各州捐款数据占比
biden_state.plot.pie(figsize=(10, 10),autopct='%0.2f%%',subplots=True)

array([<AxesSubplot:ylabel='TRANSACTION_AMT'>], dtype=object)

# 首先下载图片模型,这里提供的是已经处理好的图片，有兴趣的选手可以自己写代码进行图片处理
# 处理结果要求：需要将人图像和背景颜色分离，并纯色填充，词云才会只显示在人图像区域
# 拜登原图：https://img.alicdn.com/tfs/TB1pUcwmZVl614jSZKPXXaGjpXa-689-390.jpg
# 拜登处理后图片：https://img.alicdn.com/tfs/TB10Jx4pBBh1e4jSZFhXXcC9VXa-689-390.jpg
# 特朗普原图：https://img.alicdn.com/tfs/TB1D0l4pBBh1e4jSZFhXXcC9VXa-298-169.jpg
# 特朗普处理后图片：https://img.alicdn.com/tfs/TB1BoowmZVl614jSZKPXXaGjpXa-298-169.jpg
# 在jupyter notebook下你可以直接用下面命令下载处理后的图片
!wget https://img.alicdn.com/tfs/TB10Jx4pBBh1e4jSZFhXXcC9VXa-689-390.jpg

--2020-10-23 10:44:50--  https://img.alicdn.com/tfs/TB10Jx4pBBh1e4jSZFhXXcC9VXa-689-390.jpg
Resolving img.alicdn.com (img.alicdn.com)... 101.37.183.187, 101.37.183.186
Connecting to img.alicdn.com (img.alicdn.com)|101.37.183.187|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 4236 (4.1K) [image/jpeg]
Saving to: ‘TB10Jx4pBBh1e4jSZFhXXcC9VXa-689-390.jpg’

100%[======================================>] 4,236       --.-K/s   in 0s      

2020-10-23 10:44:50 (54.8 MB/s) - ‘TB10Jx4pBBh1e4jSZFhXXcC9VXa-689-390.jpg’ saved [4236/4236]

# 由于下载图片文件名过长，我们对文件名进行重命名
import os
os.rename('TB10Jx4pBBh1e4jSZFhXXcC9VXa-689-390.jpg', 'biden.jpg')

# 如果有兴趣，可以在上面下载原图后进行图片处理
# 拜登原图：https://img.alicdn.com/tfs/TB1pUcwmZVl614jSZKPXXaGjpXa-689-390.jpg
# 拜登处理后图片：https://img.alicdn.com/tfs/TB10Jx4pBBh1e4jSZFhXXcC9VXa-689-390.jpg
'''
欢迎大家评论区分享你的图片处理代码
'''

# 在4.2 热门候选人拜登在各州的获得的捐赠占比 中我们已经取出了所有支持拜登的人的数据，存在变量：biden中
# 将所有捐赠者姓名连接成一个字符串
data = ' '.join(biden["NAME"].tolist())
# 读取图片文件
bg = plt.imread("biden.jpg")
# 生成
wc = WordCloud(# FFFAE3
    background_color="white",  # 设置背景为白色，默认为黑色
    width=890,  # 设置图片的宽度
    height=600,  # 设置图片的高度
    mask=bg,    # 画布
    margin=10,  # 设置图片的边缘
    max_font_size=100,  # 显示的最大的字体大小
    random_state=20,  # 为每个单词返回一个PIL颜色
).generate_from_text(data)
# 图片背景
bg_color = ImageColorGenerator(bg)
# 开始画图
plt.imshow(wc.recolor(color_func=bg_color))
# 为云图去掉坐标轴
plt.axis("off")
# 画云图，显示
# 保存云图
wc.to_file("biden_wordcloud.png")

推荐阅读
 ···  END  ···