1

我遇到了一个我无法理解的奇怪的 python 错误：

import pandas as pd
import numpy as np

def logit(p):
    if 0 < p < 1:
        return np.log(p / (1 - p))
    elif p == 0 or p == 1:
        return pd.NA
    else:
        raise ValueError(f"logit input must be [0,1], got {p}")

class D(object):
    d = pd.DataFrame({'prob': np.random.rand(100)})
    transformation = logit
    def transform1(self):
       d = self.d
       self.transformation = [self.transformation]
       for trans in self.transformation:
           d['prob'] = d['prob'].apply(trans)

    @staticmethod
    def transform2(d, transformation):
       transformation = [transformation]
       for trans in transformation:
           d['prob'] = d['prob'].apply(trans)

当我做：

In [5]: dat = D()

In [6]: dat.transform1()
--------------------------------------------------------------
TypeError                    Traceback (most recent call last)
<ipython-input-6-602e90bcb76e> in <module>
----> 1 dat.transform1()

<ipython-input-4-3de477ddc9d9> in transform1(self)
     17        self.transformation = [self.transformation]
     18        for trans in self.transformation:
---> 19            d['prob'] = d['prob'].apply(trans)
     20
     21     @staticmethod

~/anaconda3/envs/py377/lib/python3.7/site-packages/pandas/core/series.py in apply(self, func, convert_dtype, args, **kwds)
   3846             else:
   3847                 values = self.astype(object).values
-> 3848                 mapped = lib.map_infer(values, f, convert=convert_dtype)
   3849
   3850         if len(mapped) and isinstance(mapped[0], Series):

pandas/_libs/lib.pyx in pandas._libs.lib.map_infer()

TypeError: logit() takes 1 positional argument but 2 were given

但其他类似的实现工作正常

In [8]: dat.transform2(dat.d, logit)

In [9]: dat.d.head()
Out[9]:
       prob
0 -0.766136
1 -0.279224
2  1.209884
3 -0.461204
4 -1.265117

似乎在transform1，当我调用applyapd.DataFrame时，类对象本身被传递到应用函数中：

In [1]: import pandas as pd
   ...: import numpy as np
   ...:
   ...: def logit(*p):
   ...:     print(p)
   ...:     if 0 < p < 1:
   ...:         return np.log(p / (1 - p))
   ...:     elif p == 0 or p == 1:
   ...:         return pd.NA
   ...:     else:
   ...:         raise ValueError(f"logit input must be [0,1],
   ...: got {p}")

In [2]: dat.transform1()
(<__main__.D object at 0x7fb2b01cda50>, 0.24971056160452076)
-------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-2-602e90bcb76e> in <module>
----> 1 dat.transform1()

为什么会这样？我正在使用Python 3.7.7. 谢谢！

4

0 回答 0