今天这里谈的函数，以后进行数据分析的时候会经常用到。

import numpy as npimport pandas as pdfrom pandas import DataFrame , Seriesfrom numpy import nan as NA
obj = Series(['c', 'a', 'd', 'a', 'a', 'b', 'b', 'c', 'c'])

uniques = obj.unique()print("obj is \n", obj)print("obj.unique is \n ", obj.unique())print("uniques.sort() is \n", uniques.sort())

print("obj.value_counts() is \n", obj.value_counts())print("pd.value_counts(obj.values,sort=False) \n", pd.value_counts(obj.values, sort=False))

mask = obj.isin(['b' , 'c'])print("obj.isin(['b','c']) \n", obj.isin(['b' , 'c']))print("mask = obj.isin(['b','c'])")print("obj[mask] is \n", obj[mask])

data= DataFrame({'Qu1':[1,3,4,3,4],        'Qu2':[2,3,1,2,3],        'Qu3':[1,5,2,4,4]}
)

print ("data is \n",data)
result = data.apply(pd.value_counts).fillna(0)

print("data.apply(pd.value_counts).fillna(0)\n ", result)print("计算一个series各值出现的频率")

print("handling the missing data \n")
string_data = Series(['aardvark','artichoke',np.nan,'avocado'])print("string_data is \n", string_data)print("string_data.isnull() \n",string_data.isnull())

print("The built-in python None value is also treated as NA in object Arrays \n")print("string_data[0]=None\n")
string_data[0]=Noneprint("string_data.isnull() \n ",string_data.isnull)

print(" NA handling methods in P143 Table 5-12")
data = Series([1,NA,3.5,NA,7])
data.dropna()print("data is \n",data)print("data.dropna() is \n", data.dropna())print("data[data.notnull()],\n",data[data.notnull()])

data = DataFrame([[1.,6.5,3.],[1.,NA,NA],[NA,NA,NA],[NA,6.5,3.]])
cleaned = data.dropna()

print("data is \n",data)print("data.dropna() is \n",cleaned)print("data.dropna(how='all') is \n", data.dropna(how='all'))print("passing how=all will only drop rows that are all NA")

data[4]=NAprint("New data is \n", data)print("data.dropna(axis=1,how='all') \n",data.dropna(axis=1,how='all'))print("按照columns drop")

df=DataFrame(np.random.randn(7,3))print("df is \n",df)
df.ix[:4,1]=NA
df.ix[:2,2]=NAprint("New df is \n",df)print("df.dropna(thresh=3)\n",df.dropna(thresh=3))

print("filling in the missing data")print("df.fillna(0) \n",df.fillna(0))print("df.fillna({1:0.5,3:-1}) \n",df.fillna({1:0.5,3:-1}))print("calling fillna with a dict you can use a different fill value for each columns")

_=df.fillna(0,inplace=True)print("_=df.fillna(0,inplace=True) \n",df)

df=DataFrame(np.random.randn(6,3))print("DataFrame(np.random.randn(6,3)) \n",df)
df.ix[2:,1] = NA
df.ix[4:,2] = NAprint("df.ix[2:,1] = NA; df.ix[4:,2] = NA \n",df )

print("df.fillna(method = 'ffill') \n", df.fillna(method = 'ffill'))

print("df.fillna(method = 'ffill',limit =2) \n",df.fillna(method='ffill',limit = 2))

data= Series([1.,NA,3.5,NA,7])print("data is \n",data)print("data.fillna(data.mean()) \n",data.fillna(data.mean()))

print("fillna function arguments on P146 Table 5-13")print("Hierarchical indexing")
data = Series(np.random.randn(10),index=[['a','a','a','b','b','b','c','c','d','d'],[1,2,3,1,2,3,1,2,2,3]])print("data is \n",data)print("a Series with multi-index")print("data.index",data.index)print("data['b'] \n",data['b'])print("data['b':'c'] \n",data['b':'c'])print("data.ix[['b','d']] \n",data.ix[['b','d']])

print("data[:,2] \n",data[:,2])print("data.unstack() \n",data.unstack())

print("data.unstack().stack() \n ",data.unstack().stack())

print("data frame")
frame = DataFrame(np.arange(12).reshape((4,3)),index=[['a','a','b','b'],[1,2,1,2]],columns=[['Ohio','Ohio','Colorado'],['Green','Red','Green']])print("frame is \n",frame)
frame.index.names =["key1","key2"]
frame.columns.names=["state","color"]print("New frame is \n",frame)

print("frame['Ohio'] \n",frame['Ohio'])print("frame.swaplevel('key1','key2') \n", frame.swaplevel('key1','key2'))

print("frame.sortlevel(1) \n",frame.sortlevel(1))print("frame.swaplevel(0,1).sortlevel(0)\n",frame.swaplevel(0,1).sortlevel(0))

print("summary statistics by level")print("frame.sum(level='key2') \n",frame.sum(level='key2'))print("frame.sum(level='color',axis=1) \n",frame.sum(level='color',axis = 1))

print("Using a DataFrame's columns")
frame = DataFrame({'a':range(7),'b':range(7,0,-1),'c':['one','one','one','two','two','two','two'],'d':[0,1,2,0,1,2,3]})print("frame is \n",frame)
frame2= frame.set_index(['c','d'])print("creating a new Dataframe using one or more its columns as the index")print("frame.set_index(['c','d']) \n",frame2)

frame.set_index(['c','d'],drop=False)print("frame.set_index(['c','d'],drop =False) \n",frame.set_index(['c','d'],drop=False))

print("reset_index does the opposite of set_index,the hierarchical index levels are moved into the columns")print("frame2.reset_index() \n",frame2.reset_index())

http://www.xuebuyuan.com/2180572.html