一、排序和排名
排序：sort_index和sort_values函数
代码示例：
print 'Series排序' x = Series(range(4), index = ['b', 'a', 'c', 'd']) print x.sort_index() # Series按索引排序 ''' a 1 b 0 c 2 d 3 ''' print x.sort_values() # Series按值排序 ''' b 0 a 1 c 2 d 3 ''' print 'DataFrame按索引排序' frame = DataFrame(numpy.arange(8).reshape((2, 4)), index = ['b', 'a'], columns = list('ABDC')) print frame ''' A B D C b 0 1 2 3 a 4 5 6 7 ''' print frame.sort_index() # 根据行索引来排序 ''' A B D C a 4 5 6 7 b 0 1 2 3 ''' print frame.sort_index(axis = 1) #根据列索引来排序 ''' A B C D b 0 1 3 2 a 4 5 7 6 ''' print frame.sort_index(axis = 1, ascending = False) # 设置降序排序 ''' D C B A b 2 3 1 0 a 6 7 5 4 ''' print 'DataFrame按列的值排序' frame = DataFrame({'b':[4, 7, -3, 2], 'a':[0, 1, 0, 1]}) print frame ''' a b 0 0 4 1 1 7 2 0 -3 3 1 2 ''' print frame.sort_values(by = 'b') # 指定b这列的值进行排序 ''' a b 2 0 -3 3 1 2 0 0 4 1 1 7 ''' print frame.sort_values(by = ['a', 'b']) #先a后b进行列的值排序 ''' a b 2 0 -3 0 0 4 3 1 2 1 1 7 '''排名：根据值的大小/出现次数来进行排名，得到一组排名值：rank函数
print 'rank：默认升序，排名值从1开始' obj = Series([4, 2, 0, 4],index = ['a','b','c','d']) # 以值从小到大来赋排名值：c:0(1) b:2(2) a:4(3) d:4(4) print obj.rank() ''' a 3.5 求平均值(4+3)/2 b 2.0 c 1.0 d 3.5 ''' print obj.rank(method = 'first') # 按出现顺序排名，不求平均值。 ''' a 3.0 b 2.0 c 1.0 d 4.0 ''' print obj.rank(ascending = False, method = 'max') # 逆序，并取排名值最大值。所以-5的rank是7 # a:4(1) d:4(2) b:2(3) c:0(4) ''' dtype: float64 a 2.0 b 3.0 c 4.0 d 2.0 ''' frame = DataFrame({'b':[4.3, 7, -3, 2], 'a':[0, 1, 0, 1], 'c':[-2, 5, 8, -2.5]}) print frame ''' a b c 0 0 4.3 -2.0 1 1 7.0 5.0 2 0 -3.0 8.0 3 1 2.0 -2.5 ''' print frame.rank(axis = 1) # 按行进行排名，默认升序 ''' a b c 0 2.0 3.0 1.0 1 1.0 3.0 2.0 2 2.0 1.0 3.0 3 2.0 3.0 1.0 '''
二、索引重复的情况
代码示例：
print '重复索引:进行两次索引'
obj = Series([0,1,2,3,4], index = ['a', 'a', 'b', 'b', 'c'])
print obj.index.is_unique # 判断是非有重复索引
# False
print obj['a'][0]
# 0
print obj.a[1]
# 1
df = DataFrame(numpy.arange(12).reshape(4, 3), index = ['a', 'a', 'b', 'b'])
print df
'''
0 1 2
a 0 1 2
a 3 4 5
b 6 7 8
b 9 10 11
'''
print df.ix['b'].ix[0] // 两次行索引
'''
0 6
1 7
2 8
Name: b, dtype: int32
'''
print df.ix['b'].ix[1]
'''
0 9
1 10
2 11
Name: b, dtype: int32
'''
三、汇总和计算描述统计
常用方法选项：
常用汇总统计函数 I：
常用汇总统计函数 II：
代码示例：
print '求和'
df = DataFrame([[1, numpy.nan], [7, 4], [numpy.nan, numpy.nan], [0, 1]],
index = ['a', 'b', 'c', 'd'],
columns = ['one', 'two'])
print df
'''
one two
a 1.0 NaN
b 7.0 4.0
c NaN NaN
d 0.0 1.0
'''
print df.sum() # 按列求和
# 排除缺失值，skipna默认值为True
'''
one 8.0
two 5.0
dtype: float64
'''
print df.sum(skipna = False)
'''
one NaN
two NaN
'''
print df.sum(axis = 1) # 按行求和
'''
a 1.0
b 11.0
c 0.0
d 1.0
dtype: float64
'''
print '求平均数'
print df.mean(axis = 1, skipna = False)
'''
a NaN
b 5.5
c NaN
d 0.5
'''
print df.mean(axis = 1)
'''
a 1.0
b 5.5
c NaN
d 0.5
'''
print '其它函数'
print df
'''
one two
a 1.0 NaN
b 7.0 4.0
c NaN NaN
d 0.0 1.0
'''
print df.idxmax() # 计算每一列最大值的索引
'''
one b
two b
'''
print df.cumsum() # 每一列的累加和
'''
one two
a 1.0 NaN
b 8.0 4.0
c NaN NaN
d 8.0 5.0
'''
print df.describe() # 对DataFrame每列计算汇总统计
'''
one two
count 3.000000 2.00000
mean 2.666667 2.50000
std 3.785939 2.12132
min 0.000000 1.00000
25% NaN NaN
50% NaN NaN
75% NaN NaN
max 7.000000 4.00000
'''
obj = Series([2,4,8,4], index = ['a', 'a', 'b', 'c'])
print obj.describe() # 对Series计算汇总统计
'''
count 4.000000
mean 4.500000
std 2.516611
min 2.000000
25% 3.500000
50% 4.000000
75% 5.000000
max 8.000000
dtype: float64
'''
四、相关系数与协方差
相关系数：相关系数是用以反映变量之间相关关系密切程度的统计指标。
协方差：从直观上来看，协方差表示的是两个变量总体误差的期望。如果两个
变量的变化趋势一致，也就是说如果其中一个大于自身的期望值时另外一个也
大于自身的期望值，那么两个变量之间的协方差就是正值；如果两个变量的变
化趋势相反，即其中一个变量大于自身的期望值时另外一个却小于自身的期望
值，那么两个变量之间的协方差就是负值cov函数计算协方差，corr函数计算相关系数。corrwith函数计算DataFrame的行/列与另一个Series/DataFrame的相关系数。
五、去重和成员出现计数
主要方法：
print '去重'
obj = Series(['c', 'a', 'd','b', 'b', 'c'])
print obj.unique()
'''
['c' 'a' 'd' 'b']
'''
print obj.value_counts()
'''
b 2
c 2
d 1
a 1
'''
print '判断元素存在'
mask = obj.isin(['b', 'c'])
print mask
'''
0 True
1 False
2 False
3 True
4 True
5 True
'''
print obj[mask] #只打印元素b和c
'''
0 c
3 b
4 b
5 c
'''
data = DataFrame({'Qu1':[1, 3, 4, 3, 4],
'Qu2':[2, 3, 1, 2, 3],
'Qu3':[1, 5, 2, 4, 4]})
print data
'''
Qu1 Qu2 Qu3
0 1 2 1
1 3 3 5
2 4 1 2
3 3 2 4
4 4 3 4
'''
print data.apply(pd.value_counts).fillna(0)
# 计算每列中各个数字出现的次数，缺失值为0
'''
Qu1 Qu2 Qu3
1 1.0 1.0 1.0
2 0.0 2.0 1.0
3 2.0 2.0 0.0
4 2.0 0.0 2.0
5 0.0 0.0 1.0
'''
print data.apply(pd.value_counts, axis = 1).fillna(0)
# 计算每行中各个数字出现的次数，缺失值为0
'''
1 2 3 4 5
0 2.0 1.0 0.0 0.0 0.0
1 0.0 0.0 2.0 0.0 1.0
2 1.0 1.0 0.0 1.0 0.0
3 0.0 1.0 1.0 1.0 0.0
4 0.0 0.0 1.0 2.0 0.0
'''
六、处理缺失数据
• NaN（Not a Number）表示浮点数和非浮点数组中的缺失数据，None也被当作NA处理。
处理缺失数据函数：
• dropna 函数：DatFrame默认丢弃任何含有缺失值的行。how参数控制行为，axis参数选择轴，thresh参数控制NaN数量的要求。
• fillna函数： inplace参数控制返回新对象还是就地修改
代码示例：
print '作为null处理的值'
string_data = Series(['a', 'b', numpy.nan, 'd'])
print string_data
'''
0 a
1 b
2 NaN
3 d
'''
print string_data.isnull()
'''
0 False
1 False
2 True
3 False
'''
string_data[0] = None
print string_data
'''
0 None
1 b
2 NaN
3 d
'''
# None也被当作NA处理
print string_data.isnull()
'''
0 True
1 False
2 True
3 False
'''
from numpy import nan as NA
print '丢弃缺失数据NaN'
data = Series([1, NA, 3.5, NA, 7])
print data.dropna()
'''
0 1.0
2 3.5
4 7.0
'''
print 'DataFrame对丢弃NA的处理'
data = DataFrame([[1., 6.5, 3.], [1., NA, NA],
[NA, NA, NA], [NA, 6.5, 3.]])
print data
'''
0 1 2
0 1.0 6.5 3.0
1 1.0 NaN NaN
2 NaN NaN NaN
3 NaN 6.5 3.0
'''
print data.dropna() # 默认只要某行有NA就全部删除
'''
0 1 2
0 1.0 6.5 3.0
'''
print data.dropna(how = 'all') # 某行全部为NA才删除
'''
0 1 2
0 1.0 6.5 3.0
1 1.0 NaN NaN
3 NaN 6.5 3.0
'''
data[0] = NA
print data.dropna(axis = 1, how = 'all') #某行有NA就全部删除
'''
1 2
0 6.5 3.0
1 NaN NaN
2 NaN NaN
3 6.5 3.0
'''
data = DataFrame(numpy.arange(21).reshape(7, 3))
data.ix[:4, 1] = NA
data.ix[:2, 2] = NA
print data
'''
0 1 2
0 0 NaN NaN
1 3 NaN NaN
2 6 NaN NaN
3 9 NaN 11.0
4 12 NaN 14.0
5 15 16.0 17.0
6 18 19.0 20.0
'''
print data.dropna(thresh = 2) # 每行至少要有2个非NA元素则删除
'''
0 1 2
3 9 NaN 11.0
4 12 NaN 14.0
5 15 16.0 17.0
6 18 19.0 20.0
'''
print '填充0'
df = DataFrame(numpy.arange(9).reshape(3, 3))
df.ix[:1, 1] = NA
df.ix[:2, 2] = NA
print df.fillna(0) # 默认inplace为False
'''
0 1 2
0 0 0.0 0.0
1 3 0.0 0.0
2 6 7.0 0.0
'''
print df
'''
0 1 2
0 0 NaN NaN
1 3 NaN NaN
2 6 7.0 NaN
'''
df.fillna(0, inplace = True) # 就地修改
print df
'''
0 1 2
0 0 0.0 0.0
1 3 0.0 0.0
2 6 7.0 0.0
'''
df = DataFrame(numpy.arange(9).reshape(3, 3))
df.ix[:1, 1] = NA
df.ix[:2, 2] = NA
print '不同行列填充不同的值'
print df.fillna({1:0.5, 2:-1}) # 第3列不存在
'''
0 1 2
0 0 0.5 -1.0
1 3 0.5 -1.0
2 6 7.0 -1.0
'''
print '不同的填充方式'
print df
'''
0 1 2
0 0 NaN NaN
1 3 NaN NaN
2 6 7.0 NaN
'''
print df.fillna(method = 'bfill') # 向前填充
'''
0 1 2
0 0 7.0 NaN
1 3 7.0 NaN
2 6 7.0 NaN
'''
print df.fillna(method = 'bfill', limit = 1) # 只可向前填充一步
'''
0 1 2
0 0 NaN NaN
1 3 7.0 NaN
2 6 7.0 NaN
'''
print '用统计数据填充'
data = Series([1, NA, 2, NA, 3])
print data.fillna(data.mean())
'''
0 1.0
1 2.0
2 2.0
3 2.0
4 3.0
'''
七、多层次化索引
对Series和DataFrame进行多层次的索引MultiIndex，通过stack与unstack进行Series和DataFrame的变换。
代码示例：
from pandas import MultiIndex print 'Series的多层次索引' data = Series(numpy.arange(8), index = [['a', 'a', 'b', 'b', 'c', 'c', 'd','d'], [1, 2, 1, 2, 1, 2, 1,2]]) print data # 两层行索引 ''' a 1 0 2 1 b 1 2 2 3 c 1 4 2 5 d 1 6 2 7 ''' print data.index ''' MultiIndex(levels=[[u'a', u'b', u'c', u'd'], [1, 2]], labels=[[0, 0, 1, 1, 2, 2, 3], [0, 1, 0, 1, 0, 1, 0]]) ''' print data.b ''' 1 2 2 3 ''' print data['b':'c'] # 闭区间 ''' b 1 2 2 3 c 1 4 2 5 ''' print data[:2] # 数组索引不区分标签 ''' a 1 0 2 1 ''' print data.unstack() #将Series转换为DataFrame ''' 1 2 a 0 1 b 2 3 c 4 5 d 6 7 ''' print data.unstack().stack() # 将DataFrame转换回Series ''' a 1 0 2 1 b 1 2 2 3 c 1 4 2 5 d 1 6 2 7 ''' print print 'DataFrame的多层次化索引' frame = DataFrame(numpy.arange(12).reshape((4, 3)), index = [['a', 'a', 'b', 'b'], [1, 2, 1, 2]], columns = [['A', 'A', 'B'], ['A1', 'A2', 'B1']]) print frame # 两层行索引和两层列索引 ''' A B A1 A2 B1 a 1 0 1 2 2 3 4 5 b 1 6 7 8 2 9 10 11 ''' print frame.index ''' MultiIndex(levels=[[u'a', u'b'], [1, 2]], labels=[[0, 0, 1, 1], [0, 1, 0, 1]]) ''' print frame.columns ''' MultiIndex(levels=[[u'A', u'B'], [u'A1', u'A2', u'B1']], labels=[[0, 0, 1], [0, 1, 2]]) ''' frame.index.names = ['key1', 'key2'] frame.columns.names = ['state', 'more'] print frame ''' state A B more A1 A2 B1 key1 key2 a 1 0 1 2 2 3 4 5 b 1 6 7 8 2 9 10 11 ''' print frame.ix['a', 1] ''' A A1 0 A2 1 B B1 2 ''' print frame.ix['a', 1]['B'] ''' more B1 2 ''' print frame.ix['a', 1]['A']['A1'] ''' 0 ''' print print '直接用MultiIndex创建层次索引结构index' print MultiIndex.from_arrays([['A', 'A', 'B'], ['Gree', 'Red', 'Green']], names = ['state', 'color']) ''' MultiIndex(levels=[[u'A', u'B'], [u'Gree', u'Green', u'Red']], labels=[[0, 0, 1], [0, 2, 1]], names=[u'state', u'color']) '''将索引层进行交换：swaplevel函数。对某个索引层进行排序：sortlevel函数
代码示例：
print '索引层交换' frame = DataFrame(numpy.arange(12).reshape((4, 3)), index = [['a', 'a', 'b', 'b'], [1, 2, 1, 2]], columns = [['A', 'A', 'B'], ['A1', 'A2', 'B1']]) frame.index.names = ['key1', 'key2'] print frame ''' A B A1 A2 B1 key1 key2 a 1 0 1 2 2 3 4 5 b 1 6 7 8 2 9 10 11 ''' frame_swapped = frame.swaplevel('key1', 'key2') # 交互索引层 print frame_swapped ''' A B A1 A2 B1 key2 key1 1 a 0 1 2 2 a 3 4 5 1 b 6 7 8 2 b 9 10 11 ''' print frame_swapped.swaplevel(0, 1) # 交换回来 ''' A B A1 A2 B1 key1 key2 a 1 0 1 2 2 3 4 5 b 1 6 7 8 2 9 10 11 ''' print print '对某个索引层进行排序' print frame.sortlevel('key2') ''' A B A1 A2 B1 key1 key2 a 1 0 1 2 b 1 6 7 8 a 2 3 4 5 b 2 9 10 11 ''' print frame.swaplevel(0, 1).sortlevel(0) ''' A B A1 A2 B1 key2 key1 1 a 0 1 2 b 6 7 8 2 a 3 4 5 b 9 10 11 '''根据某个索引层进行统计计算
代码示例：
print '根据索引层进行统计' print frame ''' A B A1 A2 B1 key1 key2 a 1 0 1 2 2 3 4 5 b 1 6 7 8 2 9 10 11 ''' print frame.sum(level = 'key2') ''' A B A1 A2 B1 key2 1 6 8 10 2 12 14 16 '''将某列转化为层次的行索引，列名为索引名，列的值为索引值：set_index函数；恢复重置行索引且恢复列：reset_index函数。
代码示例：
print '将列索引转化行层次索引' 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 ''' a b c d 0 0 7 one 0 1 1 6 one 1 2 2 5 one 2 3 3 4 two 0 4 4 3 two 1 5 5 2 two 2 6 6 1 two 3 ''' print frame.set_index(['c', 'd']) # 把c/d列索引变成行索引 ''' a b c d one 0 0 7 1 1 6 2 2 5 two 0 3 4 1 4 3 2 5 2 3 6 1 ''' print frame.set_index(['c', 'd'], drop = False) # 列依然保留 ''' a b c d c d one 0 0 7 one 0 1 1 6 one 1 2 2 5 one 2 two 0 3 4 two 0 1 4 3 two 1 2 5 2 two 2 3 6 1 two 3 ''' frame2 = frame.set_index(['c', 'd']) print frame2.reset_index() # 恢复列 ''' c d a b 0 one 0 0 7 1 one 1 1 6 2 one 2 2 5 3 two 0 3 4 4 two 1 4 3 5 two 2 5 2 6 two 3 6 1 '''
八、整数型索引值
Series/DataFrame的索引值的类型为整数时，使用数组索引会产生歧义：无法分清是数组类型索引还是字典类型索引。整数型索引的Series/DataFrame索引的方法：以iloc索引替代数组索引。
代码示例：
print '索引值为整数时的歧义'
ser = Series(numpy.arange(3))
print ser
'''
0 0
1 1
2 2
'''
try:
print ser[-1] # 这里会有歧义.
except:
print 'exception'
ser2 = Series(numpy.arange(3), index = ['a', 'b', 'c'])
print ser2[-1] # 索引值类型不是整数
# 2
ser3 = Series(range(3), index = [-5, 1, 3])
print ser3.iloc[2] # 使用iloc避免直接用[2]产生的歧义
# 2
print
print '对DataFrame使用整数索引'
frame = DataFrame(numpy.arange(6).reshape((3, 2)), index = [2, 0, 1])
print frame
'''
0 1
2 0 1
0 2 3
1 4 5
'''
print frame.iloc[0]
'''
0 4
1 5
'''
# print frame[2] 有歧义则会发生异常错误
# print frame['2'] 不存在'2'该索引
print frame.iloc[:, 1]
'''
2 1
0 3
1 5
'''
