葡萄酒数据分析

• 引入包，查看数据

import pandas as pd
import matplotlib.pyplot as plt
# 支持中文
plt.rcParams['font.sans-serif'] = ['SimHei']  # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False  # 用来正常显示负号
read_df = pd.read_csv('winequality-red.csv',sep=';')
white_df = pd.read_csv('winequality-white.csv',sep=';')
white_df.head()

read_df.head()

• 新增颜色列

import numpy as np
# 为红葡萄酒数据框创建颜色数组
color_red = np.repeat(0,read_df.shape[0])

# 为白葡萄酒数据框创建颜色数组
color_white = np.repeat(1,white_df.shape[0])

read_df['color'] = color_red
white_df['color'] = color_white

• 合并红白葡萄酒数据集

wine_df = read_df.append(white_df)
wine_df.info()
wine_df.to_csv('winequality_edited.csv',index=False)
wine_df.to_csv('winequality_edited1.csv')

• 可视化 固定酸度

# 固定酸度
wine_df['fixed acidity'].plot(kind='hist');

# 总二氧化硫
wine_df['total sulfur dioxide'].plot(kind='hist');
# pH 值
wine_df['pH'].plot(kind='hist');
# 酒精度
wine_df['alcohol'].plot(kind='hist');

某种类型的葡萄酒（红葡萄酒或白葡萄酒）是否代表更高的品质？

# 0 红色葡萄酒平均质量  5.636023   1 白色葡萄酒平均质量 5.877909
wine_df.groupby('color').mean()['quality']

• 查看中位数的值

wine_df.describe()

def dtype(x):
    if x <3.11:
        return '高'
    elif x <= 3.21:
        return '中等偏高'
    elif x <= 3.32:
        return '中'
    else:
        return '低'
wine_df['temp'] = wine_df['pH'].map(lambda x:dtype(x))

wine_df.groupby('temp').mean()['quality']

---

---

---

---

问题

1: 缺陷是否集中在某些区站 (各个区站的缺陷数量)

2: 最常出现缺陷的是哪种设备 (每种设备的缺陷数量统计)

3: 缺陷最集中的区站中最常出现缺陷的是哪种设备

4: 最常出现缺陷的设备经常出现哪种缺陷

• 引入包

import numpy as np
import pandas as pd
# % matplotlib inline
import pymongo
import matplotlib.pyplot as plt
# 支持中文
plt.rcParams['font.sans-serif'] = ['SimHei']  # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False  # 用来正常显示负号

myclient = pymongo.MongoClient("mongodb://localhost:13300/")
bj_itpsdsc = myclient.get_database('bj')
phm_analysis_proproblem = bj_itpsdsc.phm_analysis_proproblem

• 构造设备类型字典

devCodeMapDF = pd.read_excel('设备种类编码映射.xlsx')
devCodeMap={}
for i in range(devCodeMapDF.shape[0]):
    devCodeMap[devCodeMapDF.iloc[i]['种类编码']]=devCodeMapDF.iloc[i]['设备种类']

• mongodb 查出数据，根据字典映射设备类型

proproblem = phm_analysis_proproblem.find({})
proproblemDF=pd.DataFrame(list(proproblem))
proproblemDF['devCodeMap'] = proproblemDF['devCode'].map(lambda x: devCodeMap[int(x)])

• 设备种类 对应缺陷数

devCodeCounts = proproblemDF['devCodeMap'].value_counts()
devCodeCounts

import matplotlib
fig =proproblemDF['devCode'].value_counts().plot(title='缺陷设备种类分类数量直方图',kind='bar')
fig.get_figure().savefig('缺陷设备种类分类数量直方图（无中文）.png')

fig = proproblemDF['devCodeMap'].value_counts().plot(title='缺陷设备种类分类数量直方图', kind='bar',figsize=(16,6))
fig.get_figure().savefig('缺陷设备种类分类数量直方图.png')

fig = proproblemDF['devCodeMap'].value_counts().plot(title='缺陷设备种类分类饼图', kind='pie',figsize=(10,9));
fig.get_figure().savefig('缺陷设备种类分类饼图.png')

结果文档