女性服装数据分析（电商数据）版本1

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
color = sns.color_palette()

data = pd.read_csv('Womens_Clothing.csv')
#  查看数据结构
data

	Unnamed: 0	Clothing ID	Age	Title	Review Text	Rating	Recommended IND	Positive Feedback Count	Division Name	Department Name	Class Name
0	0	767	33	NaN	Absolutely wonderful - silky and sexy and comf...	4	1	0	Initmates	Intimate	Intimates
1	1	1080	34	NaN	Love this dress! it's sooo pretty. i happene...	5	1	4	General	Dresses	Dresses
2	2	1077	60	Some major design flaws	I had such high hopes for this dress and reall...	3	0	0	General	Dresses	Dresses
3	3	1049	50	My favorite buy!	I love, love, love this jumpsuit. it's fun, fl...	5	1	0	General Petite	Bottoms	Pants
4	4	847	47	Flattering shirt	This shirt is very flattering to all due to th...	5	1	6	General	Tops	Blouses
...	...	...	...	...	...	...	...	...	...	...	...
23481	23481	1104	34	Great dress for many occasions	I was very happy to snag this dress at such a ...	5	1	0	General Petite	Dresses	Dresses
23482	23482	862	48	Wish it was made of cotton	It reminds me of maternity clothes. soft, stre...	3	1	0	General Petite	Tops	Knits
23483	23483	1104	31	Cute, but see through	This fit well, but the top was very see throug...	3	0	1	General Petite	Dresses	Dresses
23484	23484	1084	28	Very cute dress, perfect for summer parties an...	I bought this dress for a wedding i have this ...	3	1	2	General	Dresses	Dresses
23485	23485	1104	52	Please make more like this one!	This dress in a lovely platinum is feminine an...	5	1	22	General Petite	Dresses	Dresses

23486 rows × 11 columns

有上面结果可知：

该数据集包括23486行和10个特征变量。每行对应一个客户评论，并包含以下变量：

**服装ID：**整数分类变量，指的是要查看的特定作品。
**年龄：**评论者年龄的正整数变量。
**标题：**评论标题的字符串变量。
**评论文本：**评论正文的字符串变量。
**评分：**客户授予的产品评分的正序整数变量，从1最差，到5最佳。
**推荐的IND：**二进制变量，说明客户在推荐1的地方推荐产品，不推荐0的地方。
**积极的反馈计数：**积极的整数，记录发现该评论为积极的其他客户的数量。
**高级部门名称：**产品高级部门的分类名称。
**部门名称：**产品部门名称的分类名称。
**类名称：**产品类名称的分类名称。

中文名称 英文名称

服装ID Clothing ID

年龄 Age

标题 Title

评论文本 Review Text

评分： Rating

推荐的IND Recommended IND

积极的反馈计数 Positive Feedback Count

高级部门名称 Division Name

部门名称 Department Name

类名称 Class Name

data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 23486 entries, 0 to 23485
Data columns (total 11 columns):
Unnamed: 0                 23486 non-null int64
Clothing ID                23486 non-null int64
Age                        23486 non-null int64
Title                      19676 non-null object
Review Text                22641 non-null object
Rating                     23486 non-null int64
Recommended IND            23486 non-null int64
Positive Feedback Count    23486 non-null int64
Division Name              23472 non-null object
Department Name            23472 non-null object
Class Name                 23472 non-null object
dtypes: int64(6), object(5)
memory usage: 2.0+ MB

#  查看缺失值
# data.isnull()
#  删除缺失值
df = data.dropna()
df

	Unnamed: 0	Clothing ID	Age	Title	Review Text	Rating	Recommended IND	Positive Feedback Count	Division Name	Department Name	Class Name
2	2	1077	60	Some major design flaws	I had such high hopes for this dress and reall...	3	0	0	General	Dresses	Dresses
3	3	1049	50	My favorite buy!	I love, love, love this jumpsuit. it's fun, fl...	5	1	0	General Petite	Bottoms	Pants
4	4	847	47	Flattering shirt	This shirt is very flattering to all due to th...	5	1	6	General	Tops	Blouses
5	5	1080	49	Not for the very petite	I love tracy reese dresses, but this one is no...	2	0	4	General	Dresses	Dresses
6	6	858	39	Cagrcoal shimmer fun	I aded this in my basket at hte last mintue to...	5	1	1	General Petite	Tops	Knits
...	...	...	...	...	...	...	...	...	...	...	...
23481	23481	1104	34	Great dress for many occasions	I was very happy to snag this dress at such a ...	5	1	0	General Petite	Dresses	Dresses
23482	23482	862	48	Wish it was made of cotton	It reminds me of maternity clothes. soft, stre...	3	1	0	General Petite	Tops	Knits
23483	23483	1104	31	Cute, but see through	This fit well, but the top was very see throug...	3	0	1	General Petite	Dresses	Dresses
23484	23484	1084	28	Very cute dress, perfect for summer parties an...	I bought this dress for a wedding i have this ...	3	1	2	General	Dresses	Dresses
23485	23485	1104	52	Please make more like this one!	This dress in a lovely platinum is feminine an...	5	1	22	General Petite	Dresses	Dresses

19662 rows × 11 columns

分析

# 1. 可视化 给出评分者的年龄
plt.hist(df['Age'], color=color[1], label='age')
plt.legend()
plt.xlabel('age')
plt.ylabel('count')
plt.title('age of commentator')
print('\n figure 01')

 figure 01

得出结论

由figure01 可得出：给出评论的人的年龄大多在25到45之间，青年、中年人较多

# 2. 可视化不同年龄的等级图
plt.figure(figsize=(10, 8))
sns.boxplot(x='Rating', y='Age', data=df)
plt.title('age of rating')
print('\n figure 02')

 figure 02

得出结论

由figure02 可得出：给出评分分布的年龄都差不多

3、每个部门、推荐什么服装？
查看Division Name,Department Name和’Class Name的唯一值

print('高级部门Division Name', df['Division Name'].unique())
print()
print('部门Department Name',df['Department Name'].unique())
print()
print('类名称Class Name',df['Class Name'].unique())

高级部门Division Name ['General' 'General Petite' 'Initmates']部门Department Name ['Dresses' 'Bottoms' 'Tops' 'Intimate' 'Jackets' 'Trend']类名称Class Name ['Dresses' 'Pants' 'Blouses' 'Knits' 'Intimates' 'Outerwear' 'Lounge''Sweaters' 'Skirts' 'Fine gauge' 'Sleep' 'Jackets' 'Swim' 'Trend' 'Jeans''Shorts' 'Legwear' 'Layering' 'Casual bottoms' 'Chemises']

将Recommended IND推荐产品为1，不推荐0的数据分开

# recommend  not_recommend
recommend = df[df['Recommended IND'] == 1]
not_recommend = df[df['Recommended IND'] == 0]
# recommend.head()
not_recommend.head()

	Unnamed: 0	Clothing ID	Age	Title	Review Text	Rating	Recommended IND	Positive Feedback Count	Division Name	Department Name	Class Name
2	2	1077	60	Some major design flaws	I had such high hopes for this dress and reall...	3	0	0	General	Dresses	Dresses
5	5	1080	49	Not for the very petite	I love tracy reese dresses, but this one is no...	2	0	4	General	Dresses	Dresses
10	10	1077	53	Dress looks like it's made of cheap material	Dress runs small esp where the zipper area run...	3	0	14	General	Dresses	Dresses
22	22	1077	31	Not what it looks like	First of all, this is not pullover styling. th...	2	0	7	General	Dresses	Dresses
25	25	697	31	Falls flat	Loved the material, but i didnt really look at...	3	0	0	Initmates	Intimate	Lounge

# 4.可视化不同部门的推荐和不推荐的叠加柱状图
plt.figure(figsize=(12,8))
plt.hist(recommend['Department Name'], color=color[2], alpha=0.5, label='recommend')
plt.hist(not_recommend['Department Name'], color=color[4], alpha=0.5, label='not_recommend')
plt.legend()
plt.xticks(rotation=45)
plt.title('Department recommend and not_recommend')
print('\n figure 03')

 figure 03

得出结论

由figure03可知 绿色的面积大于X色的面积，由此说明，大部分部门都可以推荐商品

# 可视化不同商品的推荐和不推荐叠加柱状图
plt.figure(figsize=(12,8))
plt.hist(recommend['Class Name'], color=color[1], alpha=0.5, label='recommend')
plt.hist(not_recommend['Class Name'], color=color[5], alpha=0.5, label='not_recommend')
plt.legend()
plt.xticks(rotation=45)
plt.title('Class recommend and not_recommend')
print('\n figure 04')

 figure 04

得出结论

从figure04看出：并不是卖最多的Knits商品推荐成功率最大

# 哪个年龄段的人对什么样的衣服发表什么样的评论
df['Review Length'] = df['Review Text'].astype(str).apply(len)
df

E:\anaconda\lib\site-packages\ipykernel_launcher.py:2: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value insteadSee the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

​

	Unnamed: 0	Clothing ID	Age	Title	Review Text	Rating	Recommended IND	Positive Feedback Count	Division Name	Department Name	Class Name	Review Length
2	2	1077	60	Some major design flaws	I had such high hopes for this dress and reall...	3	0	0	General	Dresses	Dresses	500
3	3	1049	50	My favorite buy!	I love, love, love this jumpsuit. it's fun, fl...	5	1	0	General Petite	Bottoms	Pants	124
4	4	847	47	Flattering shirt	This shirt is very flattering to all due to th...	5	1	6	General	Tops	Blouses	192
5	5	1080	49	Not for the very petite	I love tracy reese dresses, but this one is no...	2	0	4	General	Dresses	Dresses	488
6	6	858	39	Cagrcoal shimmer fun	I aded this in my basket at hte last mintue to...	5	1	1	General Petite	Tops	Knits	496
...	...	...	...	...	...	...	...	...	...	...	...	...
23481	23481	1104	34	Great dress for many occasions	I was very happy to snag this dress at such a ...	5	1	0	General Petite	Dresses	Dresses	131
23482	23482	862	48	Wish it was made of cotton	It reminds me of maternity clothes. soft, stre...	3	1	0	General Petite	Tops	Knits	223
23483	23483	1104	31	Cute, but see through	This fit well, but the top was very see throug...	3	0	1	General Petite	Dresses	Dresses	208
23484	23484	1084	28	Very cute dress, perfect for summer parties an...	I bought this dress for a wedding i have this ...	3	1	2	General	Dresses	Dresses	427
23485	23485	1104	52	Please make more like this one!	This dress in a lovely platinum is feminine an...	5	1	22	General Petite	Dresses	Dresses	110

19662 rows × 12 columns

#  绘制单Review Length变量分布
# 单变量分布的最方便的方法是sns.distplot()功能。默认情况下，这将绘制直方图并拟合核密度估计（KDE）
fig = plt.figure(figsize=(12, 8))
ax = sns.distplot(df['Review Length'], color=color[3])
ax = plt.title("Length of Reviews")
print('\n figure 05')

 figure 05

得出结论

由figure05可得出 大部分人评论的长度都基本在500

#  可视化不同年龄段的评论长度分布
plt.figure(figsize=(18,8))
sns.boxplot(x='Age', y='Review Length', data=df)
print('\n figure 06')

 figure 06

# 评分与正面反馈计数
plt.figure(figsize=(12,8))
sns.boxplot(x = 'Rating', y = 'Positive Feedback Count', data = df)
print('\n figure 07')

 figure 07

得出结论

由图figure07可得出 评分在3以上的正面反馈的计数大

词云评论可视化

# 1. 数据清洗
import re
from wordcloud import WordCloud, STOPWORDSdef clean_data(text):letters_only = re.sub("[^a-zA-Z]", " ", text) #  替换标点符合等words = letters_only.lower().split()                            return( " ".join( words ))
#     return letters_onlystopwords= set(STOPWORDS)|{'skirt', 'blouse','dress','sweater', 'shirt','bottom', 'pant', 'pants' 'jean', 'jeans','jacket', 'top', 'dresse'}def create_cloud(rating):x= [i for i in rating]y= ' '.join(x)cloud = WordCloud(background_color='white',width=1600, height=800,max_words=100,stopwords= stopwords).generate(y)plt.figure(figsize=(15,7.5))plt.axis('off')plt.imshow(cloud)plt.show()

#  等级是5的词云图
rating5= df[df['Rating']==5]['Review Text'].apply(clean_data)
create_cloud(rating5)

#  等级是4的词云图
rating4= df[df['Rating']==4]['Review Text'].apply(clean_data)
create_cloud(rating4)

#  等级是3的词云图
rating3= df[df['Rating']==3]['Review Text'].apply(clean_data)
create_cloud(rating3)

#  等级是2的词云图
rating2= df[df['Rating']==2]['Review Text'].apply(clean_data)
create_cloud(rating2)

#  等级是1的词云图
rating1= df[df['Rating']==1]['Review Text'].apply(clean_data)
create_cloud(rating1)