应用Pandas创建带有函数的列方法

Question

我正在尝试优化代码并节省时间。我当前的解决方案有效，但当我将类似的函数应用于多个数据帧时，它是冗余的，无法维护。

如何根据其他列的条件自动创建新列？

部分数据：

import pandas as pd

df = {'Column1': [1,2,3,4,5],
        'Column2': ["A","B","C","D","E"]}
df = pd.DataFrame(df, columns=['Column1','Column2'])

df


Column1 Column2
0   1   A
1   2   B
2   3   C
3   4   D
4   5   E

方法1：每次我需要使用新数据帧执行类似操作时工作，但不可维护

# create band if column 2 contains A-C
df['Col_2_Band V1'] = "D-E"
df['Col_2_Band V1'][df['Column2'].isin(['A','B','C'])] = "A-C"
df


Column1 Column2 Col_2_Band V1
0   1   A   A-C
1   2   B   A-C
2   3   C   A-C
3   4   D   D-E
4   5   E   D-E

方法2：无法开始工作

def applyV2(row):
    row['Col_2_Band V2'] = "D-E"
    row['Col_2_Band V2'][df['Column2'].isin(['A','B','C'])] = "A-C"
    return row

df = df.apply(applyV2, axis=1)

**Error:**
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-8-cf5d31427d02> in <module>()
      4     return row
      5 
----> 6 df = df.apply(applyV2, axis=1)

C:\Users\cfeld\AppData\Local\Continuum\Anaconda3\lib\site-packages\pandas\core\frame.py in apply(self, func, axis, broadcast, raw, reduce, args, **kwds)
   4852                         f, axis,
   4853                         reduce=reduce,
-> 4854                         ignore_failures=ignore_failures)
   4855             else:
   4856                 return self._apply_broadcast(f, axis)

C:\Users\cfeld\AppData\Local\Continuum\Anaconda3\lib\site-packages\pandas\core\frame.py in _apply_standard(self, func, axis, ignore_failures, reduce)
   4948             try:
   4949                 for i, v in enumerate(series_gen):
-> 4950                     results[i] = func(v)
   4951                     keys.append(v.name)
   4952             except Exception as e:

<ipython-input-8-cf5d31427d02> in applyV2(row)
      1 def applyV2(row):
      2     row['Col_2_Band V2'] = "D-E"
----> 3     row['Col_2_Band V2'][df['Column2'].isin(['A','B','C'])] = "A-C"
      4     return row
      5 

TypeError: ("'str' object does not support item assignment", 'occurred at index 0')

结束目标：将此方法应用于多个dfs

# for example

df_10 = df10.apply(applyV2, axis=1)
df_20 = df20.apply(applyV2, axis=1)
df_30 = df30.apply(applyV2, axis=1)

Answer 1

如果可能，请不使用pd.DataFrame.apply来处理易于执行的功能。 df.apply只是一个薄薄的循环。

在这种情况下，以下更有效，同样可维护。 pd.DataFrame.pipe只是将数据帧放入函数中。我们使用.loc访问器来分配取决于给定条件的值。

def add_row(df):
    df['Col_2_Band V2'] = 'D-E'
    df.loc[df['Column2'].isin({'A','B','C'}), 'Col_2_Band V2'] = 'A-C'
    return df

df = df.pipe(add_row)

Answer 2

它不是最干净的，但你可以这样做：

import pandas as pd
df = {'Column1': [1,2,3,4,5],
  'Column2': ["A","B","C","D","E"]}
df = pd.DataFrame(df, columns=['Column1','Column2'])

def applyV2(x):
    df['Col_2_Band v2'] = df['Column2'].map(lambda x: "A-C" if "A" in x
                                                       else 'A-C' if 'B' in x
                                                       else 'A-C' if 'C' in x
                                                       else 'D-E')
    return x
df.apply(applyV2)

输出：

Column1 Column2 Col_2_Band v2
0   1      A        A-C
1   2      B        A-C
2   3      C        A-C
3   4      D        D-E
4   5      E        D-E