Python导入plydata库失败，出现TypeError错误求助

我安装了plydata 0.4.3，但导入该库时出现错误，报错信息如下：

TypeError                                 Traceback (most recent call last)
Cell In[7], line 2
      1 import sys
----> 2 import plydata

File ~\anaconda3\Lib\site-packages\plydata\__init__.py:1
----> 1 from .one_table_verbs import *  # noqa
      2 from .two_table_verbs import *  # noqa
      3 from .helper_verbs import *     # noqa

File ~\anaconda3\Lib\site-packages\plydata\one_table_verbs.py:6
      1 """
      2 One table verb initializations
      3 """
      4 import itertools
----> 6 from .operators import DataOperator
      7 from .expressions import Expression
      9 __all__ = ['define', 'create', 'sample_n', 'sample_frac', 'select',
     10            'rename', 'distinct', 'unique', 'arrange', 'group_by',
     11            'ungroup', 'group_indices', 'summarize',
     12            'filter', 'count', 'add_count', 'add_tally',
     13            'summarise', 'mutate', 'transmute',
     14            ]

File ~\anaconda3\Lib\site-packages\plydata\operators.py:6
      2 from collections import defaultdict
      4 import pandas as pd
----> 6 from .eval import EvalEnvironment
      7 from .types import GroupedDataFrame
      8 from .utils import temporary_attr, custom_dict

File ~\anaconda3\Lib\site-packages\plydata\eval.py:22
     18             flags |= feature.compiler_flag
     19     return flags
---> 22 _ALL_FUTURE_FLAGS = _all_future_flags()
     25 @lru_cache(maxsize=256)
     26 def _compile(source, filename, mode, flags=0, dont_inherit=False, optimize=-1):
     27     """
     28     Cached compile
     29     """

File ~\anaconda3\Lib\site-packages\plydata\eval.py:17, in _all_future_flags()
     15 for feature_name in __future__.all_feature_names:
     16     feature = getattr(__future__, feature_name)
---> 17     if feature.getMandatoryRelease() > sys.version_info:
     18         flags |= feature.compiler_flag
     19 return flags

TypeError: '>' not supported between instances of 'NoneType' and 'sys.version_info'

我的Python版本是3.11，相关包信息如下：

Name: plydata
Version: 0.4.3
Summary: Functions for manipulating Data in Python
Home-page: https://github.com/has2k1/plydata
Author: 
Author-email: 
License: BSD (3-clause)
Location: C:\Users\marco\anaconda3\Lib\site-packages
Requires: pandas
Required-by: 
Note: you may need to restart the kernel to use updated packages.

-----------------------------

Name: pandas
Version: 1.5.3
Summary: Powerful data structures for data analysis, time series, and statistics
Home-page: https://pandas.pydata.org
Author: The Pandas Development Team
Author-email: pandas-dev@python.org
License: BSD-3-Clause
Location: C:\Users\marco\anaconda3\Lib\site-packages
Requires: numpy, numpy, python-dateutil, pytz
Required-by: bokeh, datashader, holoviews, hvplot, pandas-datareader,
panel, plydata, pmdarima, seaborn, statsmodels, xarray, yahoo-fin
Note: you may need to restart the kernel to use updated packages.

问题原因 #

plydata 0.4.3的eval.py文件中，_all_future_flags函数未处理feature.getMandatoryRelease()返回None的情况。在Python 3.11及以上版本中，部分__future__特性的强制发布版本会返回None，直接与sys.version_info比较就会触发类型错误。

解决方案 #

• 方案1：升级plydata
  执行以下命令升级到兼容Python 3.11的版本（推荐0.5.0及以上）：

  pip install --upgrade plydata
  

• 方案2：手动修改源码
  找到本地anaconda环境中plydata的eval.py文件（路径参考：C:\Users\marco\anaconda3\Lib\site-packages\plydata\eval.py），修改_all_future_flags函数，添加对None的判断：

  def _all_future_flags():
      flags = 0
      for feature_name in __future__.all_feature_names:
          feature = getattr(__future__, feature_name)
          release = feature.getMandatoryRelease()
          if release is not None and release > sys.version_info:
              flags |= feature.compiler_flag
      return flags
  

  修改后保存文件，重新导入plydata即可。

内容的提问来源于stack exchange，提问作者Alien