MCPcopy Create free account
hub / github.com/apache/arrow / read_row_group

Method read_row_group

python/pyarrow/parquet/core.py:445–490  ·  view source on GitHub ↗

Read a single row group from a Parquet file. Parameters ---------- i : int Index of the individual row group that we want to read. columns : list If not None, only these columns will be read from the row group. A column na

(self, i, columns=None, use_threads=True,
                       use_pandas_metadata=False)

Source from the content-addressed store, hash-verified

443 return self.reader.closed
444
445 def read_row_group(self, i, columns=None, use_threads=True,
446 use_pandas_metadata=False):
447 """
448 Read a single row group from a Parquet file.
449
450 Parameters
451 ----------
452 i : int
453 Index of the individual row group that we want to read.
454 columns : list
455 If not None, only these columns will be read from the row group. A
456 column name may be a prefix of a nested field, e.g. 'a' will select
457 'a.b', 'a.c', and 'a.d.e'.
458 use_threads : bool, default True
459 Perform multi-threaded column reads.
460 use_pandas_metadata : bool, default False
461 If True and file has custom pandas schema metadata, ensure that
462 index columns are also loaded.
463
464 Returns
465 -------
466 pyarrow.table.Table
467 Content of the row group as a table (of columns)
468
469 Examples
470 --------
471 >>> import pyarrow as pa
472 >>> table = pa.table({'n_legs': [2, 2, 4, 4, 5, 100],
473 ... 'animal': ["Flamingo", "Parrot", "Dog", "Horse",
474 ... "Brittle stars", "Centipede"]})
475 >>> import pyarrow.parquet as pq
476 >>> pq.write_table(table, 'example.parquet')
477 >>> parquet_file = pq.ParquetFile('example.parquet')
478
479 >>> parquet_file.read_row_group(0)
480 pyarrow.Table
481 n_legs: int64
482 animal: string
483 ----
484 n_legs: [[2,2,4,4,5,100]]
485 animal: [["Flamingo","Parrot",...,"Brittle stars","Centipede"]]
486 """
487 column_indices = self._get_column_indices(
488 columns, use_pandas_metadata=use_pandas_metadata)
489 return self.reader.read_row_group(i, column_indices=column_indices,
490 use_threads=use_threads)
491
492 def read_row_groups(self, row_groups, columns=None, use_threads=True,
493 use_pandas_metadata=False):

Calls 1

_get_column_indicesMethod · 0.95