无法使用 pd.read_parquet 读取镶木地板

Question

我刚刚更新了所有 conda 环境（pandas 1.4.1），并且面临 pandas read_parquet 函数的问题。

parquet_file = r'F:\Python Scripts\my_file.parquet'
file= pd.read_parquet(path = parquet_file)

它会生成以下错误：

file= pd.read_parquet(parquet_file)
Traceback (most recent call last):

  File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\api.py:135 in _parse_header
    fmd = read_thrift(f, parquet_thrift.FileMetaData)

  File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\thrift_structures.py:25 in read_thrift
    obj.read(pin)

  File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\parquet_thrift\parquet\ttypes.py:1929 in read
    iprot._fast_decode(self, iprot, [self.__class__, self.thrift_spec])

TypeError: got wrong ttype while reading field


During handling of the above exception, another exception occurred:

Traceback (most recent call last):

  Input In [14] in <cell line: 1>
    whole_df42= pd.read_parquet(parquet_file)

  File C:\ProgramData\Anaconda3\lib\site-packages\pandas\io\parquet.py:493 in read_parquet
    return impl.read(

  File C:\ProgramData\Anaconda3\lib\site-packages\pandas\io\parquet.py:345 in read
    parquet_file = self.api.ParquetFile(path, **parquet_kwargs)

  File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\api.py:100 in __init__
    self._parse_header(fn, verify)

  File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\api.py:138 in _parse_header
    self.fn)

AttributeError: 'ParquetFile' object has no attribute 'fn'

我尝试过不同的镶木地板文件，但没有任何结果。

我尝试重新创建镶木地板，只需使用：

df.to_parquet(parquet_file)

from a DataFrame 但又出现了另一个错误。

      Traceback (most recent call last):
      File F:\Python Scripts\My_modules\io\df.py:86 in dataframe_to_parquet
        df.to_parquet(full_path, **kwargs)
    
      File C:\ProgramData\Anaconda3\lib\site-packages\pandas\util\_decorators.py:207 in wrapper
        return func(*args, **kwargs)
    
      File C:\ProgramData\Anaconda3\lib\site-packages\pandas\core\frame.py:2835 in to_parquet
        return to_parquet(
    
      File C:\ProgramData\Anaconda3\lib\site-packages\pandas\io\parquet.py:420 in to_parquet
        impl.write(
    
      File C:\ProgramData\Anaconda3\lib\site-packages\pandas\io\parquet.py:301 in write
        self.api.write(
    
      File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\writer.py:938 in write
        write_simple(filename, data, fmd, row_group_offsets,
    
      File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\writer.py:802 in write_simple
        rg = make_row_group(f, data[start:end], fmd.schema,
    
      File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\writer.py:671 in make_row_group
        chunk = write_column(f, data[column.name], column,
    
      File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\writer.py:589 in write_column
        bdata = compress_data(bdata, compression)
    
      File C:\ProgramData\Anaconda3\lib\site-packages\fastparquet\compression.py:113 in compress_data
        raise RuntimeError("Compression '%s' not available.  Options: %s" %
    
    RuntimeError: Compression 'snappy' not available.  Options: ['BROTLI', 'GZIP', 'UNCOMPRESSED']

我已经使用了这些代码行数十次，我认为这是与新版本有关的问题。有人有解决方法或可以帮助我了解问题所在吗？

谢谢。