pandas Panel 的使用
1、可以使用多种方式创建Panel，3D ndarrayimport pandas as pdimport numpy as npdata = np.random.rand(2,4,5)p = pd.Panel(data)print p
2、DataFrame Objects的dict代码如下：import pandas as pdimport numpy as ndata = {'num1' : pd.DataFrame(np.random.randn(2, 3)), 'num2' : pd.DataFrame(np.random.randn(5, 7))}p = pd.Panel(data)print p输出结果：<class 'pandas.core.panel.Panel'>Dimensions: 2 (items) x 5 (major_axis) x 7 (minor_axis)Items axis: num1 to num2Major_axis axis: 0 to 4Minor_axis axis: 0 to 6Process finished with exit code 0
3、从面板中选择数据可以使用方法panel.major_axis（index）访问数据。import pandas as pdimport numpy as npdata = {'num1' : pd.DataFrame(np.random.randn(2, 3)), 'num2' : pd.DataFrame(np.random.randn(5, 7))}p = pd.Panel(data)print p.major_xs(1)输出： num1 num20 -0.482575 0.6441231 -0.586697 0.9262722 -0.458282 1.6554563 NaN -0.2248774 NaN -0.5153185 NaN -0.8022146 NaN -0.449112
4、创建一个Series并查看所有上面列表的属性操作import pandas as pdimport numpy as nps = pd.Series(np.random.randn(19))print s输出：Connected to pydev debugger (build 181.5087.37)0 1.2968861 0.3741762 0.4570023 0.3097324 -0.3604745 -0.4007946 0.395560
5、轴返回系列标签的列表。import pandas as pdimport numpy as nps = pd.Series(np.random.randn(19))print ("The axes are:")print s.axes输出结果：Connected to pydev debugger (build 181.5087.37)The axes are:[RangeIndex(start=0, stop=19, step=1)]
6、NDIM返回对象的维数。根据定义，Series是一维数据结构import pandas as pdimport numpy as nps = pd.Series(np.random.randn(19))print sprint ("The dimensions of the object:")print s.ndim输出：1
7、形状返回表示DataFrame维度的元组。元组（a，b），其中a表示行墙绅褡孛数，b表示列数。import pandas as pdd = {'Name': pd.Series(['zhangsan', 'xiaomin', 'lisi', 'lili']), 'Phone': pd.Series([13537106722, 13537106721, 13537106726, 13737106726]), 'Age': pd.Series([22, 24, 25 , 26])}# Create a DataFramedf = pd.DataFrame(d)print ("object:")print dfprint ("shape:")print df.shape输出：object: Age Name Phone0 22 zhangsan 135371067221 24 xiaomin 135371067212 25 lisi 135371067263 26 lili 13737106726shape:(4, 3)
