pandas--数据处理

s = pd.Series([1,3,6,np.NaN,44,1])

print(s)

dates = pd.date_range('20181203',periods=6)
print(dates)

dates = pd.date_range('20181203',periods=7)
date_df = pd.DataFrame(np.random.randn(7,4),index=dates,columns=['a','b','c','d'])
print(date_df)

df1 = pd.DataFrame(np.arange(12).reshape(3,4))
print(df1)

df = pd.DataFrame({'A':1,'B':[1,2,3],'C':'hello'})
print(df)

df = pd.DataFrame({'A':1,'B':[1,2,3],'C':'hello'})
print(df.dtypes)
print(df.index)
print(df.columns)
print(df.values)

df = pd.DataFrame({'A':1,'B':[1,2,3],'C':'hello'})
print(df)
df = df.sort_index(axis=1,ascending=False)
print(df)

df = pd.DataFrame({'A':1,'B':[1,2,3],'C':'hello'})
print(df)
df = df.sort_values(by='B',ascending=False)
print(df)

dates = pd.date_range('20180101',periods=6)
print(dates)
df = pd.DataFrame(np.arange(24).reshape(6,4),index=dates,columns=['A','B','C','D'])
print(df)

#按列选择元素
print(df.A,df['A'])

#按行选择元素

print(df[0:3],df['20180103':'20180105'])

#按照lable 选择元素,既可以按照index又可以按照列属性
print(df.loc['20180102'])
print(df.loc[:,['B']])

#select by position
print(df.iloc[1])
print(df.iloc[:,1])
print(df.iloc[1:3,1:3])

#mixed selection,既可以按照标签，又可以按照position
print(df.ix[:,'B'])

#boolean indexing
print(df.A>8)
print(df[df.A>8])
print(df.loc[df.A>8])

dates = pd.date_range('20180101',periods=6)
df = pd.DataFrame(np.arange(24).reshape(6,4),index=dates,columns=['A','B','C','D'])
df.A[df.A>12] = 5
print(df)

dates = pd.date_range('20180101',periods=6)
df = pd.DataFrame(np.arange(24).reshape(6,4),index=dates,columns=['A','B','C','D'])
df[df.A>12] = 5
print(df)

dates = pd.date_range('20180101',periods=6)
df = pd.DataFrame(np.arange(24).reshape(6,4),index=dates,columns=['A','B','C','D'])
df.iloc[0,1] = np.NaN
df.iloc[1,2] = np.NaN
print(df)
#按行丢掉，how={'any','all'} 'all' 为只有该行全部为nan才丢掉该行
print(df.dropna(axis=0,how='any'))

dates = pd.date_range('20180101',periods=6)
df = pd.DataFrame(np.arange(24).reshape(6,4),index=dates,columns=['A','B','C','D'])
df.iloc[0,1] = np.NaN
df.iloc[1,2] = np.NaN
print(df)
col_mean = df.mean()
#mean 为：该列的和/该列不为nan的总数
print(col_mean)
print(df.fillna(value=col_mean))

print(np.any(df.isnull())==True)

df 在fillna 后一定要重新赋值

df = df.fillna(value=col_mean)