@@ -244,9 +244,10 @@ def to_parquet(df, path, engine='auto', compression='snappy', **kwargs):
244
244
path : string
245
245
File path
246
246
engine : {'auto', 'pyarrow', 'fastparquet'}, default 'auto'
247
- Parquet reader library to use. If 'auto', then the option
248
- 'io.parquet.engine' is used. If 'auto', then the first
249
- library to be installed is used.
247
+ Parquet library to use. If 'auto', then the option
248
+ ``io.parquet.engine`` is used. The default ``io.parquet.engine``
249
+ behavior is to try 'pyarrow', falling back to 'fastparquet' if
250
+ 'pyarrow' is unavailable.
250
251
compression : {'snappy', 'gzip', 'brotli', None}, default 'snappy'
251
252
Name of the compression to use. Use ``None`` for no compression.
252
253
kwargs
@@ -271,9 +272,10 @@ def read_parquet(path, engine='auto', columns=None, **kwargs):
271
272
272
273
.. versionadded 0.21.1
273
274
engine : {'auto', 'pyarrow', 'fastparquet'}, default 'auto'
274
- Parquet reader library to use. If 'auto', then the option
275
- 'io.parquet.engine' is used. If 'auto', then the first
276
- library to be installed is used.
275
+ Parquet library to use. If 'auto', then the option
276
+ ``io.parquet.engine`` is used. The default ``io.parquet.engine``
277
+ behavior is to try 'pyarrow', falling back to 'fastparquet' if
278
+ 'pyarrow' is unavailable.
277
279
kwargs are passed to the engine
278
280
279
281
Returns
0 commit comments