我的Python分析成长之路9

 1 import pandas as pd
 2 ser1 = pd.Series([1,2,3,4])
 3 print(ser1)
 4 print(ser1.index)  #ser1的索引
 5 print(ser1.values)  #ser1的值
 6 ser2= pd.Series([1,2,3,4],index=['a','b','c','d'])  #自己设置索引
 7 print(ser2['a'])  #获得索引为a的值
 8 print(ser2[['a','b','c']])#获取多个索引值\
 9 #Series对象自身和其索引都有name属性，
10 ser2.name ="p"
11 ser2.index.name = 'state'
12 print(ser2)

 1 import pandas as pd
 2 import numpy as np
 3 data ={'state':['a','a','a','b','b','b'],
 4        "year":[2000,2001,2002,2001,2002,2003],
 5        "pop":[1.5,1.7,3.6,2.4,2.9,3.2]}
 6 df1 = pd.DataFrame(data)
 7 print(df1)
 8 df2 = pd.DataFrame(data,columns=['year','state','pop'],index=["one","two","three","four","five","six"]) #自己设置列名
 9 print(df2.columns)
10 print(df2['year'])  #获取year列值
11 print(df2.year)
12 print(df2.loc["one"])  #获取one行值
13 df2['debt'] = np.arange(6)
14 print(df2)

 1 import pandas as pd
 2 import numpy as np
 3 data ={'state':['a','a','a','b','b','b'],
 4        "year":[2000,2001,2002,2001,2002,2003],
 5        "pop":[1.5,1.7,3.6,2.4,2.9,3.2]}
 6 df2 = pd.DataFrame(data,columns=['year','state','pop'],
 7                    index=["one","two","three","four","five","six"])
 8 print(df2['year'])   #从DataFrame中选择单列或列序列
 9 print(df2.loc["one"])  #从DataFrame中选择单行或多行
10 print(df2.loc[:,"year"])  #从DataFrame中选择单列
11 print(df2.loc["one","year"])   #同时确定行和列
12 print(df2.loc["one",['year','state']]) #一行两列
13 print(df2.loc["condition",[]])  #loc 可以接受表达式
14 print(df2.iloc[1])  #选取单行
15 print(df2.iloc[:,[2,3]])  #选取索引为第2,3列
16 # print(df2.ilov[where_i,where_j][condition])
17 df2.iloc[condition,[]].values  #iloc方法不能接受表达式，条件返回的是一个Series,取出Series的值

1 import numpy as np
2 import pandas as pd
3 df4 = pd.DataFrame(np.arange(8).reshape(2,4),index=["three","one"],columns=["d","a","b","c"])
4 print(df4.sort_index(axis=1,ascending=False)) #反序
5 print(df4.sort_values(by='b')) #根据b列数值排序
6 print(df4.sort_values(by=['a','b']))
7 df5 = pd.Series([7,-5,7,4,2,0,4])
8 print(df5.rank())
9 print(df5.rank(method='first')) #按照值在数据中出现的次序分配排名

 1  1 import pandas_datareader.data as web
 2  2 tickers =['AAPL',"IBM","GOOG"]
 3  3 all_data = {ticker:web.get_data_yahoo(ticker) for ticker in tickers}
 4  4 price = pd.DataFrame({ticker:data['Adj Close'] for ticker,data in all_data.items()})
 5  5 volums = pd.DataFrame({ticker:data['Volume'] for ticker,data in all_data.items()})
 6  6 returns = price.pct_change()
 7  7 print(returns.tail())
 8  8 print(returns['AAPL'].corr(returns['IBM']))
 9  9 print(returns.corr())   #计算相关性
10 10 print(returns.cov())  #计算协整性
11 11 print(returns.corrwith(volums))

 1 import pandas as pd
 2 df = pd.DataFrame({'key1':["a","a","b","b","a"],
 3               "key2":["one","two","one","two","one"],
 4               "data1":np.random.randn(5),
 5               "data2":np.random.randn(5)})
 6 print(df)
 7 # group = df['data1'].groupby(df['key1']) #对data1列进行分组，获得一个group对象
 8 # group= df.groupby(df['key1'])['data1']   #两个等价
 9 group = df.groupby(df['key1'])  #对整个DataFrame分组
10 print(group.count())  #返回分组的数目
11 print(group.head())  #返回每组的前几个值
12 print(group.max())   #返回每组的最大值
13 print(group.mean())  #返回每组的均值
14 print(group.median())  #返回每组的中位数
15 print(group.cumcount())  #对每个分组中的成员进行标记
16 print(group.size())  #返回每个分组的大小
17 print(group.min())  #返回每个分组的最小值
18 print(group.std())  #返回每组的标准差
19 print(group.sum())  #返回每组的和
20 group2 = df['data1'].groupby([df['key1'],df['key2']])  #根据key1,key2分组

 1 import pandas as pd
 2 df = pd.DataFrame({'key1':["a","a","b","b","a"],
 3               "key2":["one","two","one","two","one"],
 4               "data1":np.random.randn(5),
 5               "data2":np.random.randn(5)})
 6 group = df['data1'].groupby(df['key1']) #对data1列进行分组，获得一个group对象
 7 group= df.groupby(df['key1'])['data1']   #两个等价
 8 group = df.groupby(df['key1'])  #对整个DataFrame分组
 9 print(group.count())  #返回分组的数目
10 print(group.head())  #返回每组的前几个值
11 print(group.max())   #返回每组的最大值
12 print(group.mean())  #返回每组的均值
13 print(group.median())  #返回每组的中位数
14 print(group.cumcount())  #对每个分组中的成员进行标记
15 print(group.size())  #返回每个分组的大小
16 print(group.min())  #返回每个分组的最小值
17 print(group.std())  #返回每组的标准差
18 print(group.sum())  #返回每组的和
19 print(group.quantile(0.9))  #返回每组的分位数
20 group2 = df['data1'].groupby([df['key1'],df['key2']])  #根据key1,key2分组
21 print(group.agg(np.mean))  #返回均值
22 def f(x):
23     return x.max()-x.min()
24 print(group.agg(f))      #使用自定义函数
25 group3 = df.groupby(df["key1"])
26 print(group3.agg({"data1":np.mean,"data2":np.max}))  #对data1 和 data2分别操作

import pandas as pd
df = pd.DataFrame({'key1':["a","a","b","b","a"],
              "key2":["one","two","one","two","one"],
              "data1":np.random.randn(5),
              "data2":np.random.randn(5)})
group = df['data1'].groupby(df['key1']) #对data1列进行分组，获得一个group对象
# group= df.groupby(df['key1'])['data1']   #两个等价
group = df.groupby(df['key1'])  #对整个DataFrame分组
print(group.count())  #返回分组的数目
print(group.head())  #返回每组的前几个值
print(group.max())   #返回每组的最大值
print(group.mean())  #返回每组的均值
print(group.median())  #返回每组的中位数
print(group.cumcount())  #对每个分组中的成员进行标记
print(group.size())  #返回每个分组的大小
print(group.min())  #返回每个分组的最小值
print(group.std())  #返回每组的标准差
print(group.sum())  #返回每组的和
print(group.quantile(0.9))  #返回每组的分位数
group2 = df['data1'].groupby([df['key1'],df['key2']])  #根据key1,key2分组
print(group.agg(np.mean))  #返回均值
def f(x):
    return x.max()-x.min()
print(group.agg(f))      #使用自定义函数
group3 = df.groupby(df["key1"])
print(group3.agg({"data1":np.mean,"data2":np.max}))  #对data1 和 data2分别操作
print(group3[["data1","data2"]].apply(lambda x:(x.mean()-x.min())/(x.max()-x.min())))#使用自定义函数

l = pd.pivot_table(df[["key1",'data1',"data2"]],index="key1")
print（l)

l2 = pd.crosstab(index=df["key1"],columns=df["key2"],values=df["data1"],aggfunc=np.sum)
print(l2)