pandas是python为数据分析建造的可靠工具，很多地方和R语言有想通之处。数据分析并不是工具越高深越好，excel,R,python都是针对不同情况的不同工具，各有各的优缺点， 就像你要搭一个架子，或者做一个工艺品，有的小锤子比较合适，有的就得用大斧子了。 excel其实是数据分析的强力武器。对于小数据量的情况下，excel有其先天的优势。而Python和R更像一个高性能的数据处理工具。 然而仅仅会使用各种厉害工具，数据落不了地是啥也没用的。落地，和业务贴合，永远是数据最终走向。近期转到研发部门，和业务贴合的机会少了，对这方面更有很多深刻感触。 下面这几点可能更像从excel角度去看python. 1.panda 的 index 更像横坐标 x，同时也可以把它当做一个list ，可以像数组一样赋值，取数。 2.两个跟属性判断相关的语句。in 是判断某个columns 或者 index中是否存在某个字段。 is 是判断数据格式类型。 3.另外基本功能，reindex 调整 index的工具函数。 基本格式：

obj.reindex(['a','b','c','d','e'],fill_value = 0.0) 功能：可以在横坐标和纵坐标上进行修改

frame = DataFrame(np.arange(9).reshape(3,3),index=['a','b','c'],columns=['Ohio','Texas','California']) frame.reindex(['a','b','c','d']) frame.reindex(columns=['Ohio','Texas','California','NewYork'])

4.取数据子矩阵 frame.ix 函数，同时还可以有reindex的功能。

frame.ix(['a','b','d'],states)

data.ix[['Colorado','Utah'],['three','four']]

5.现在已经隐约可以感觉到 python的两个子模块运算类似于线性代数了 所以 两个dataframe的结果十分像线性代数的结果。 df1+df2

df1.add(df2,fill_value=0) df1.mul(df2,fill_value=0) df1.div(df2,fill_value=0) df1.sub(df2,fill_value=0)

详细数据如下 ======================================================================================== '''panda's index objects are responsible for holding the axis labels,like series''' import pandas as pd obj = Series(range(3),index=['a','b','c']) index = obj.index index index[1:] '''index = immutable''' index[1]='d' '''so the index can be valued by function''' index = pd.Index(np.arange(3)) obj2 = Series([1.5,-2.5,0],index=index) obj2 ''' evaluate the attribute of index 判断属性用Is，判断存不存在用in''' obj2.index is index 'Ohio' in frame3.columns '2002' in obj2.index '''Essential functionality''' '''reindexing''' obj=Series([4.5,7.2,-5.3,3.6],index=['d','b','a','c']) obj2=obj.reindex(['a','b','c','d','e']) obj2 '''fill the missing data''' obj.reindex(['a','b','c','d','e'],fill_value = 0.0) '''ordering fill the missing data''' obj3=Series(['blue','green','black'],index=[0,2,4]) obj3.reindex(np.arange(5),method='ffill') '''reindex can be alter row,column and both in data frame''' frame = DataFrame(np.arange(9).reshape(3,3),index=['a','b','c'],columns=['Ohio','Texas','California']) frame.reindex(['a','b','c','d']) frame.reindex(columns=['Ohio','Texas','California','NewYork']) months = ['APR','MAY','JUN','JUL','AUG'] frame.reindex(columns=months) label=['a','b','c','d','e'] states=['Ohio','Texas','California','NewYork'] '''reindex 仅对x-axis有效''' frame.reindex(label,method='ffill') '''取子矩阵''' frame.ix(['a','b','d'],states) '''dropping entries from axis''' obj = Series(np.arange(5.),index=['a','b','c','d','e']) new_obj = obj.drop('c') new_obj '''drop from data frame''' data=DataFrame(np.arange(16).reshape(4,4),index=['Ohio','Colorado','Utah','NewYork'],columns=['one','two','three','four']) '''drop from index''' data.drop(['Colorado','Utah']) '''drop from column''' data.drop('two',axis=1) '''index,selection,filtering''' obj=Series(np.arange(4.),index=['a','b','c','d']) '''index可以像数组一样，通过数字定位,index 定位,取一个数，一串数''' obj['b'] obj[1] obj[1:2] obj[['a','c','d']] obj[[1,3]] obj[obj < 2] obj['b':'c']=5 data=DataFrame(np.arange(16).reshape(4,4),index=['Ohio','Colorado','Utah','New York'],columns=['one','two','three','four']) '''follow by columns,但只是单维度的''' data['two'] data[['three','one']] data.ix['Ohio'] data[data['three']>5] data[:2] '''把data小于5的赋值0''' data[data<5]=0 '''按照位置选择值''' data.ix['Colorado','two'] data.ix['Colorado',['two','three']] data.ix[['Colorado','Utah'],['three','four']] data.ix[2] data.ix[:'Utah','two'] data.ix[:2,'two'] data.ix[data.three>5,:3] '''reindex''' data.ix[['Colorado','Utah'],[3,0,1]] '''arithmetic and data alignment''' s1=Series([7.3,-2.5,3.4,1.5],index=['a','c','d','e']) s2=Series([-2.1,3.6,-1.5,4,3.1],index=['a','c','e','f','g']) '''not overlap return NA''' s1+s2 '''dataframe''' df1=DataFrame(np.arange(9.).reshape(3,3),columns=list('bcd'),index=['Ohio','Texas','Colorado']) df2=DataFrame(np.arange(12.).reshape(4,3),columns=list('bde'),index=['Utah','Ohio','Texas','Oregon']) df1+df2 '''只要有一个为空，就是空''' df1.add(df2,fill_value=0) '''只要有一个有数，另外一个就设为0''' '''reindex''' df1.reindex(columns=df2.columns,fill_value=0) df1 = DataFrame(np.arange(12.).reshape(3,4),columns=list('abcd')) df2 = DataFrame(np.arange(20.).reshape(4,5),columns=list('abcde')) df1.add(df2,fill_value=0) df1.mul(df2,fill_value=0) df1.div(df2,fill_value=0) df1.sub(df2,fill_value=0)