pandas - 如何在 groupby 计算（）之后保留 Dask DataFrame

Question

我有一个 Dask DataFrame，当我有一个 groupby 时，我发现在使用 compute() 之前我没有处理列，但是在使用 compute() 时 Dask DataFrame 更改为 Pandas DataFrame，所以 Dask DataFrame 没有优势，我想保留 Dask一直是DataFrame，查看详情：</p>

import pandas as pd
import dask.dataframe as dd
df = pd.DataFrame({"name":["Jack","Jack","Willom","Willom","James","James","Morgan"],
                   "fix_num":[50,50,70,70,90,90,100],
                   "score1":[50,60,70,80,90,40,60],
                   "score2":[90,50,30,40,100,80,80]})
ddf = dd.from_pandas(df, npartitions=1)

ddf.compute()

     name  fix_num  score1  score2
0    Jack       50      50      90
1    Jack       50      60      50
2  Willom       70      70      30
3  Willom       70      80      40
4   James       90      90     100
5   James       90      40      80
6  Morgan      100      60      80

def _element_coment(t):
    a = t["score1"].sum()
    b = t["score2"].sum()
    return pd.Series((a, b), index=['sum_1', 'sum_2'])


grp = ddf.groupby(['name','fix_num'])\
         .apply(_element_coment,meta={'sum_1':int, 'sum_2':int})\
         .reset_index()

judg = grp.fix_num <= grp.sum_2

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "D:\ProgramData\Anaconda3\lib\site-packages\dask\dataframe\core.py", line 3387, in __getattr__
    raise AttributeError("'DataFrame' object has no attribute %r" % key)
AttributeError: 'DataFrame' object has no attribute 'fix_num'

grp.columns        #I found no fix_num in columns

Index(['index', 'sum_1', 'sum_2'], dtype='object')

grp_2 = grp.compute()    
grp_2

     name  fix_num  sum_1  sum_2
0    Jack       50    110    140
1   James       90    130    180
2  Morgan      100     60     80
3  Willom       70    150     70

# grp_2 have fix_num in columns, but grp_2 is pandas DataFrame
jud g = grp2_2.fix_num<=grp2_2.sum_2  
grp_2.dtypes

name       object
fix_num     int64
sum_1       int64
sum_2       int64
dtype: object**

那么如何保留 Dask DataFrame 进行处理呢？

标签： pandasgroup-bydask