这篇教程MySQL数据优化-多层索引写得很实用，希望能帮到您。
一、多层索引
1.创建环境：Jupyter import numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['上半年','上半年','下半年','下半年'], ['一季度','二季度','三季度','四季度']], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','白菜','牛肉','猪肉']])display(a) 
2.设置索引的名称import numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['上半年','上半年','下半年','下半年'], ['一季度','二季度','三季度','四季度']], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','白菜','牛肉','猪肉']])a.index.names=['年度','季度']a.columns.names=['大类','小类']display(a) 
3.from_arrays( )-from_tuples()import numpy as npimport pandas as pdindex=pd.MultiIndex.from_arrays([['上半年','上半年','下半年','下半年'],['一季度','二季度','三季度','四季度']])columns=pd.MultiIndex.from_tuples([('蔬菜','胡萝卜'),('蔬菜','白菜'),('肉类','牛肉'),('肉类','猪肉')])a=pd.DataFrame(np.random.random(size=(4,4)),index=index,columns=columns)display(a)
4.笛卡儿积方式from_product() 局限性较大 import pandas as pdindex = pd.MultiIndex.from_product([['上半年','下半年'],['蔬菜','肉类']])a=pd.DataFrame(np.random.random(size=(4,4)),index=index)display(a) 
二、多层索引操作
1.Seriesimport pandas as pda=pd.Series([1,2,3,4],index=[['a','a','b','b'],['c','d','e','f']])print(a)print('---------------------')print(a.loc['a'])print('---------------------')print(a.loc['a','c'])
import pandas as pda=pd.Series([1,2,3,4],index=[['a','a','b','b'],['c','d','e','f']])print(a)print('---------------------')print(a.iloc[0])print('---------------------')print(a.loc['a':'b'])print('---------------------')print(a.iloc[0:2])
2.DataFrameimport numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['上半年','上半年','下半年','下半年'], ['一季度','二季度','三季度','四季度']], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','白菜','牛肉','猪肉']])print(a)print('--------------------')print(a.loc['上半年','二季度'])print('--------------------')print(a.iloc[0])
3.交换索引swaplevel( )
import numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['2021','2021','2022','2022'], ['一季度','二季度','三季度','四季度']], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','白菜','牛肉','猪肉']])a.index.names=['年度','季度']print(a)print('--------------------')print(a.swaplevel('年度','季度'))
4.索引排序sort_index( )
level：指定根据哪一层进行排序，默认为最层inplace：是否修改原数据。默认为False
import numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['2021','2021','2022','2022'], [1,3,2,4]], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','白菜','牛肉','猪肉']])a.index.names=['年度','季度']print(a)print('--------------------')print(a.sort_index())print('--------------------')print(a.sort_index(level=1))
5.索引堆叠stack( )
将指定层级的列转换成行 import numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['2021','2021','2022','2022'], [1,3,2,4]], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','胡萝卜','牛肉','牛肉']])print(a)print('--------------------')print(a.stack(0))print('--------------------')print(a.stack(-1))
6.取消堆叠unstack( )
将指定层级的行转换成列 fill_value：指定填充值。
import numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['2021','2021','2022','2022'], [1,3,2,4]], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','胡萝卜','牛肉','牛肉']])print(a)print('--------------------')a=a.stack(0)print(a)print('--------------------')print(a.unstack(-1))
import numpy as npimport pandas as pda=pd.DataFrame(np.random.random(size=(4,4)),index=[['2021','2021','2022','2022'], [1,3,2,4]], columns=[['蔬菜','蔬菜','肉类','肉类'],['胡萝卜','胡萝卜','牛肉','牛肉']])print(a)print('--------------------')a=a.stack(0)print(a)print('--------------------')print(a.unstack(0,fill_value='0'))
到此这篇关于MySQL数据优化-多层索引的文章就介绍到这了,更多相关数据优化-多层索引内容请搜索51zixue.net以前的文章或继续浏览下面的相关文章希望大家以后多多支持51zixue.net！
Python JSON模块的使用详情
Python制作基础学生信息管理系统 |
