Question

我有两个数据基:df1, 形状(34,1151649),df2, 形状(76, 3467)。我愿执行<条码>pandas。它们相当缓慢。我不禁要问,有什么办法加快该守则。样本代码如下。

df1 = pd.DataFrame(np.arange(6).reshape(2, 3), index=[1, 2], columns=pd.MultiIndex.from_tuples((( a , 1), ( a , 2), ( b , 3)), names=[ n1 ,  n2 ]))
df2 = pd.DataFrame(np.arange(6).reshape(3, 2), index=[0, 1, 2], columns=pd.Index([ a ,  c ], name= n1 ))

df1
df2
df1.where(df2 == 2).groupby(level=0, axis=1).sum()

产出如下:

Answer 1

<><>Example>

import pandas as pd
import numpy as np
np.random.seed(1)
df1 = pd.DataFrame(np.random.randint(0, 10, (100, 1000000)), columns=pd.MultiIndex.from_product([[ a ,  b ,  c ,  d ], range(0, 250000)], names=[ n1 ,  n2 ]))
df2 = pd.DataFrame(np.arange(6).reshape(3, 2), columns=pd.Index([ a ,  c ], name= n1 ))

df:

100 X 1000000 sample.

df2 is same with you

Code

df1.where(df2 == 2).groupby(level=0, axis=1).sum()  --> too slow

df1.groupby(level=0, axis=1).sum().where(df2 == 2, 0) --> fast

友情链接