day2+学习笔记+7/17Python-pandas之数据整合 import numpy as np import pandas as pd pd.concat([df1,df2,df3]) pd.concat([df1,df2,df3],keys=["x","y","z"]) result3 = pd.concat([result,df4],axis=1) result4 = pd.concat([result,df4],axis=1,join="inner")#完全匹配才被保留 result5 = pd.concat([result,df4],axis=1,join_axes=[result.index])
append来做concatenationdf1.append([df2,df3])
s1 = pd.Series([60,50],index["Shangha","Beijing"],name="meal") pd.concat([df1,s1],axis=1) s2 = pd.Series([18000,12000],index=["apts","cars],name="Xiamen") df1.append(s2) Merge(Join)result = result.reset_index().rename(columns={"index":"cities"}) df4.reset_index().rename(columns={"index":"cities"}) pd.merge(result,df4,on="cities")#inner join pd.merge(result,df4,on="cities",how="left")#"right"标记join方式 join on indexdf4 = df4.set_index("cities") df1.join(df4,how="right")#"left","outer" pd.merge(df1,df4,left_index=True,right_index=True,how="outer") pd.concat([df1,df4],axis=1,join_axes=[df1.index])
goog = pd.read_csv("data/GOOG.csv",index_col=0,parse_datas=["Date"]) goog.shape
@matplotlib inline goog["Adj close"].plot() app1 = pd.read_csv("data/AAPL.csv",index_col=0,parse_datas=["Date"]) app1.shape app1["adj close"].plot()#出现了null值 不能被转换成浮点型 app1["adj close"] [app1["adj close"] =="null"] = np.NaN app1["adj close"] = app1["adj close"].ffill() #前向填充 app1["adj close"].apply(lambda x:float(x)) app1["adj close"].plot()
stocks = pd.concat([app1[adj close],d[adj close],f[adj close]],axis=1,keys=["app1","d","f"]) stocks.plot() 前面的值不要了 valid_stocks = stocks[stocks.index > stocks["goog"].first_valid_index()] valid_stocks.plot(grid=True)
|