pyarrow.feather.read_feather(source, columns=None, use_threads=True, memory_map=True)[source]#

Read a pandas.DataFrame from Feather format. To read as pyarrow.Table use feather.read_table.

sourcestr file path, or file-like object
columnssequence, optional

Only read a specific set of columns. If not provided, all columns are read.

use_threadsbool, default True

Whether to parallelize reading using multiple threads. If false the restriction is used in the conversion to Pandas as well as in the reading from Feather format.

memory_mapbool, default True

Use memory mapping when opening file on disk