DataFrame對(duì)象的創(chuàng)建，修改，合并

import pandas as pdimport numpy as np

創(chuàng)建DataFrame對(duì)象

# 創(chuàng)建DataFrame對(duì)象df = pd.DataFrame([1, 2, 3, 4, 5], columns=['cols'], index=['a','b','c','d','e'])print df

 colsa 1b 2c 3d 4e 5

df2 = pd.DataFrame([[1, 2, 3],[4, 5, 6]], columns=['col1','col2','col3'], index=['a','b'])print df2

 col1 col2 col3a 1 2 3b 4 5 6

df3 = pd.DataFrame(np.array([[1,2],[3,4]]), columns=['col1','col2'], index=['a','b'])print df3

col1 col2a 1 2b 3 4

df4 = pd.DataFrame({'col1':[1,3],'col2':[2,4]},index=['a','b'])print df4

 col1 col2a 1 2b 3 4

創(chuàng)建DataFrame對(duì)象的數(shù)據(jù)可以為列表，數(shù)組和字典，列名和索引為列表對(duì)象

基本操作

# DataFrame對(duì)象的基本操作df2.index

Index([u'a', u'b'], dtype='object')

df2.columns

Index([u'col1', u'col2', u'col3'], dtype='object')

# 根據(jù)索引查看數(shù)據(jù)df2.loc['a'] # 索引為ａ這一行的數(shù)據(jù)# df2.iloc[0] 跟上面的操作等價(jià)，一個(gè)是根據(jù)索引名，一個(gè)是根據(jù)數(shù)字索引訪問(wèn)數(shù)據(jù)

col1 1col2 2col3 3Name: a, dtype: int64

print df2.loc[['a','b']] # 訪問(wèn)多行數(shù)據(jù)，索引參數(shù)為一個(gè)列表對(duì)象

col1 col2 col3a 1 2 3b 4 5 6

print df.loc[df.index[1:3]]

colsb 2c 3

# 訪問(wèn)列數(shù)據(jù)print df2[['col1','col3']]

col1 col3a 1 3b 4 6

計(jì)算

# DataFrame元素求和# 默認(rèn)是對(duì)每列元素求和print df2.sum()

col1 5col2 7col3 9dtype: int64

# 行求和print df2.sum(1)

a 6b 15dtype: int64

# 對(duì)每個(gè)元素乘以２print df2.apply(lambda x:x*2)

col1 col2 col3a 2 4 6b 8 10 12

# 對(duì)每個(gè)元素求平方(支持ndarray一樣的向量化操作)print df2**2

 col1 col2 col3a 1 4 9b 16 25 36

列擴(kuò)充# 對(duì)DataFrame對(duì)象進(jìn)行列擴(kuò)充df2['col4'] = ['cnn','rnn']print df2

 col1 col2 col3 col4a 1 2 3 cnnb 4 5 6 rnn

# 也可以通過(guò)一個(gè)新的DataFrame對(duì)象來(lái)定義一個(gè)新列,索引自動(dòng)對(duì)應(yīng)df2['col5'] = pd.DataFrame(['MachineLearning','DeepLearning'],index=['a','b'])print df2

 col1 col2 col3 col4  col5a 1 2 3 cnn MachineLearningb 4 5 6 rnn DeepLearning