Python pandas中多个聚合和列重命名的语法

请有人解释一下为什么这个agg（）不适合我：

import pandas as pd 

df = pd.DataFrame({'name':["alice","bob","charlene","alice","bob","charlene","alice","bob","charlene","edna" ],
                   'date':["2020-01-01","2020-01-01","2020-01-01","2020-01-01","2020-01-01","2020-01-01","2020-01-02","2020-01-01","2020-01-02","2020-01-01"],
                   'contribution': [5,5,10,20,30,1,5,5,10,100],
                   'payment-type': ["cash","transfer","cash","transfer","cash","transfer","cash","transfer","cash","transfer",]})
df['date'] = pd.to_datetime(df['date'])

daily_count = df.groupby(pd.Grouper(key='date', freq='1D')).agg({'name': 'value_counts', 'contribution': 'sum'}).rename(columns={'name': 'name_count', 'contribution': 'contribution_sum' }).reset_index()

我看到以下错误：

Traceback (most recent call last):
  File "/Users/andrew/git-analysis/some_pandas_test copy.py", line 11, in <module>
    daily_count = df.groupby(pd.Grouper(key='date', freq='1D')).agg({'name': ['value_counts'], 'contribution': ['sum']}).rename(columns={'name': 'name_count', 'contribution': 'contribution_sum' })
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/groupby/generic.py", line 928, in aggregate
    result, how = self._aggregate(func, *args, **kwargs)
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/base.py", line 443, in _aggregate
    return concat([result[k] for k in keys], keys=keys, axis=1), True
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/reshape/concat.py", line 271, in concat
    op = _Concatenator(
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/reshape/concat.py", line 452, in __init__
    self.new_axes = self._get_new_axes()
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/reshape/concat.py", line 515, in _get_new_axes
    return [
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/reshape/concat.py", line 516, in <listcomp>
    self._get_concat_axis() if i == self.axis else self._get_comb_axis(i)
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/reshape/concat.py", line 522, in _get_comb_axis
    return get_objs_combined_axis(
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/indexes/api.py", line 90, in get_objs_combined_axis
    return _get_combined_index(obs_idxes, intersect=intersect, sort=sort)
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/indexes/api.py", line 138, in _get_combined_index
    index = union_indexes(indexes, sort=sort)
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/indexes/api.py", line 205, in union_indexes
    result = result.union(other)
  File "/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/indexes/multi.py", line 3211, in union
    uniq_tuples = lib.fast_unique_multiple(
  File "pandas/_libs/lib.pyx", line 229, in pandas._libs.lib.fast_unique_multiple
ValueError: cannot include dtype 'M' in a buffer

回溯（最近一次呼叫最后一次）：
文件“/Users/andrew/git analysis/some_pandas_test copy.py”，第11行，在
daily_count=df.groupby（pd.Grouper（key='date'，freq='1D'））.agg（{'name'：['value_counts']，'contribution'：['sum']}）。重命名（列={'name'：'name_count'，'contribution'：'contribution_sum'}）
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/groupby/generic.py”，共928行
结果，how=self.\u聚合（func、*args、**kwargs）
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/base.py”，第443行，合计
返回concat（[result[k]表示k个键，键=键，轴=1），True
concat中的文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/reformate/concat.py”，第271行
op=\u串联器(
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/reformate/concat.py”，第452行，在__
self.new\u axes=self.\u获取\u new\u axes（）
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/reformate/concat.py”，第515行，在“获取新的”轴中
返回[
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/reformate/concat.py”，第516行，在
如果i==self.axis else self.\u获取comb\u轴（i），则获取comb\u轴（）
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site-packages/pandas/core/reformate/concat.py”，第522行，在“get\u comb\u轴”中
返回get_objs_组合轴(
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/index/api.py”，第90行，在get_objs_combined_轴中
return\u get\u combined\u索引（obs\u idx，intersect=intersect，sort=sort）
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/indexes/api.py”，第138行，在“get”和“combined”索引中
索引=联合索引（索引，排序=排序）
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/index/api.py”，第205行，联合索引中
结果=结果.联合（其他）
文件“/Library/Frameworks/Python.framework/Versions/3.8/lib/python3.8/site packages/pandas/core/index/multi.py”，第3211行，以union格式
uniq\u tuples=lib.fast\u unique\u multiple(
文件“pandas/_libs/lib.pyx”，第229行，在pandas._libs.lib.fast_unique_multiple中
ValueError:不能在缓冲区中包含数据类型“M”

如果有人知道一本好的agg（）食谱，我会很高兴看到它。我发现很难掌握这个函数

---

Ta，安德鲁

我相信这正是你想要的：

>>> (df
     .groupby(df['date'])
     .agg({'name': 'count', 'contribution': np.sum})
     .rename(columns={'name': 'name_count', 'contribution': 'contribution_sum' })
     .reset_index())
        date  name_count  contribution_sum
0 2020-01-01           8               176
1 2020-01-02           2                15

---

你有没有看过中关于聚合的部分？当我第一次了解这些东西时，它对我帮助很大

关于您的错误，您可以分别查看这两个聚合

In [6]: df.resample('1D', on='date')['name'].value_counts()
Out[6]: 
date        name    
2020-01-01  bob         3
            alice       2
            charlene    2
            edna        1
2020-01-02  alice       1
            charlene    1
Name: name, dtype: int64
In [7]: df.resample('1D', on='date')['contribution'].sum()
Out[7]: 
date
2020-01-01    176
2020-01-02     15
Freq: D, Name: contribution, dtype: int64

如果您所做的只是在日期列中按频率分组，

resample

函数比使用

Grouper

更简单。我只在需要按某个内容分组并同时重新采样时使用后者

---

如您所见（并在评论中提到）

value\u counts

返回一个多索引为

date

和

name

的序列，而

sum

只返回一个索引为

date

的序列

agg

大致是一个简短的符号，用于执行多个聚合，并将结果连接到一个数据帧，每个结果有一列。由于索引不匹配，因此无法组合这两个结果。 如果我在pandas 1.1上尝试你的代码，我会得到一个

NotImplementedError: Can only union MultiIndex with MultiIndex or Index of tuples, try mi.to_flat_index().union(other) instead.

---

基本上说，没有办法将具有不同索引级别数的序列串联起来。

value\u counts

返回一个序列，

sum

返回一个数字。我相信熊猫很难将它们对齐。您预期的结果是什么？