python.数据分析.pandas

import pandas as pd
1.用列表生成Series
s=pd.Series([1,3,5,np.nan,6,8])
print(s)

2.用Series,字典,对象生成DataFrame
df2=pd.DataFrame({'A':1.,
                  'B':pd.Timestamp('20130102'),
                  'C':pd.Series(1,index=list(range(4)),dtype='float32'),
                  'D':np.array([3]*4,dtype='int32'),
                  'E':pd.Categorical(["test","train","test","train"]),
                  'F':'foo'

})
print(df2)

3.用含日期时间索引,标签,Numpy数组生成DataFrame
d=pd.date_range('2013-01-01',periods=6)
df=pd.DataFrame(np.random.randn(6,4),index=d,columns=list('ABCD'))
print(df)
print(df.dtypes)#查看数据类型
print(df.head(2))#查看头部数据
print(df.tail(2))#查看尾部数据
print(df.columns)#查看列表名
print(df.describe())#查看数据的统计摘要
print(df.sort_index(axis=1,ascending=False))#按轴排序,降序
print(df.sort_values(by='B'))#按值排序(默认升值)
print(df['A'])#打印某一列
print(df[1:3])#切片打印
print(df['2013-01-02':'2013-01-03'])#求一行元素的数据
print(df.loc['2013-01-02'])
print(df.loc[:,['A','B']])#用标签选择多列数据
print(df.loc['2013-01-01':'2013-01-03',['A','B','C']])#标签切片
print(df.loc['2013-01-01','A'])#取出某个标签的数值
print(df.at['2013-01-01','A'])
print(df.iloc[3])#按照位置来取，取出第三行元素
print(df.iloc[1:2,2:3])#切片也可以
print(df.iloc[[1,2,3],[0,2,3]])#取出2，3，4行，1，3，4列
print(df.iloc[1:3,:])#整行切片
print(df.iloc[:,1:3])#整列切片
print(df.iloc[1,1])#提取某一个数值
print(df.iat[1,1])
print(df[df['B']>0])#帅选B中大于0的元素
print((df[df>0]))#筛选所有大于0的元素
用isin()筛选
df3=df.copy()
df3['E']=['one','one','two','three','four','three']
# print(df3)
print(df3[df3['E'].isin(['two','four'])])
用索引自动对齐新增列的数据
s1=pd.Series([1,2,3,4,5,6],index=d)
print(s1)
df['F']=s1
print(df)
按标签赋值
df.at['2013-01-01','A']=0
print(df)
#按位置赋值
df.iat[0,2]=0
print(df)
#按数组赋值
df['A']=[1,2,3,4,5,6]
print(df)
where 条件赋值
df3=df.copy()
df3[df3<0]=0
print(df3)
空值
df1=df.reindex(index=d[0:4],columns=list(df.columns)+['E'])
df1.loc[d[0]:d[1],'E']=1
print(df1)
print(df1.dropna(how='any'))#删掉空值的行
print(df1.fillna(value=5))#将空值赋值为5
#判断是否存在空值
print(pd.isna(df1))
算数运算
df1=pd.DataFrame(np.random.randn(4,5))
df2=pd.DataFrame(np.random.randn(3,4))
print(df1+df2)
#比较大小
print("df1=df2",df1.eq(df2))
print("df1!=df2",df1.ne(df2))
print("df1>df2",df1.gt(df2))
print("df1<df2",df1.lt(df2))
print("df1>=df2",df1.ge(df2))
print("df1<=df2",df1.le(df2))
统计
print(df1.mean())#求均值
print(df1.cumsum())#累加
print(pd.concat([df1,df2]))#合并

l=pd.DataFrame({'key':['foo','foo','bar'],'rval':[3,4,5]})
r=pd.DataFrame({'key':['foo','foo','bar'],'rval':[0,1,2]})
print(pd.merge(l,r,on='key'))

追加
df=pd.DataFrame(np.random.randn(8,4),columns=['A','B','C','D'])
print(df)
s=df.iloc[3]
print(s)
print(df._append(s,ignore_index=True))

分组
df=pd.DataFrame({'A':['foo','bar','foo','bar',
                      'foo','bar','foo','bar'],
                 'B':['one','two','three','four',
                      'one','two','three','four'],
                 'C':np.random.randn(8),
                 'D':np.random.randn(8)
                 })
print(df)
print(df.groupby('A').sum())
print(df.groupby(['A','B']).sum())

时间序列
rng=pd.date_range('1/1/2012',periods=100,freq='S')
ts=pd.Series(np.random.randint(0,500,len(rng)),index=rng)
print(ts)
print(ts.resample('5Min').sum())
时区表示
rng=pd.date_range('3/6/2012 00:00',periods=5,freq='D')
ts=pd.Series(np.random.randn(len(rng)),rng)
print(ts)
print(ts.plot)