Pandas.to_csv 给出错误 'ascii' codec can't encode character u'\u2013' in position 8: ordinal not in range(128)

Question

我正在尝试将熊猫数据框保存到 csv，但失败并出现错误：

df.to_csv(location, sep='|', index=False, header=True)

'ascii' 编解码器无法对位置 8 中的字符 u'\u2013' 进行编码：序数不在范围内（128）

我有熊猫版本：

>>> import pandas as pd
>>> pd.__version__
u'0.19.2'
>>>

在另一台机器上，同样的命令可以工作。安装的 pandas 版本是 0.18.1

>>> import pandas as pd
>>> pd.__version__
u'0.18.1'
>>>

我知道添加 encoding='utf-8' 可以帮助我解决错误。 但是我想知道最近是否有导致更高版本的熊猫失败的变化。

谢谢，

Answer 1

来自https://github.com/pandas-dev/pandas/blob/v0.18.1/pandas/core/frame.py 我们发现：

formatter = fmt.CSVFormatter(self, path_or_buf,
                                     line_terminator      = line_terminator,
                                     sep                  = sep,
                                     encoding             = encoding,
                                     compression          = compression,
                                     quoting              = quoting,
                                     na_rep               = na_rep,
                                     float_format         = float_format,
                                     cols                 = columns,
                                     header               = header,
                                     index                = index,
                                     index_label          = index_label,
                                     mode                 = mode,
                                     chunksize            = chunksize,
                                     quotechar            = quotechar,
                                     engine               = kwds.get("engine"),
                                     tupleize_cols        = tupleize_cols,
                                     date_format          = date_format,
                                     doublequote          = doublequote,
                                     escapechar           = escapechar,
                                     decimal              = decimal     ) 

我们发现：https://github.com/pandas-dev/pandas/blob/v0.19.2/pandas/core/frame.py

formatter = fmt.CSVFormatter(self,  path_or_buf,
                                    line_terminator =line_terminator,
                                    sep             =sep,
                                    encoding        =encoding,
                                    compression     =compression,
                                    quoting         =quoting,
                                    na_rep          =na_rep,
                                    float_format    =float_format,
                                    cols            =columns,
                                    header          =header,
                                    index           =index,
                                    index_label     =index_label,
                                    mode            =mode,
                                    chunksize       =chunksize,
                                    quotechar       =quotechar,
                                    
                                    tupleize_cols   =tupleize_cols,
                                    date_format     =date_format,
                                    doublequote     =doublequote,
                                    escapechar      =escapechar,
                                    decimal         =decimal)                                     

唯一的区别是“引擎”参数......我们现在应该更深入地研究这个“引擎”参数:-( 这里某处：https://github.com/pandas-dev/pandas/blob/v0.18.1/pandas/formats/format.py

这里： https://github.com/pandas-dev/pandas/blob/v0.19.2/pandas/formats/format.py

祝你好运！