Remove nthreads argument where thread pool is being used. Set default for use_threads to True

Change-Id: Ibf8a850f2b6889f9e4758f7063330e4fa8e324d1

python/pyarrow/feather.py

@@ -26,7 +26,6 @@ from pyarrow.compat import pdapi
 from pyarrow.lib import FeatherError  # noqa
 from pyarrow.lib import RecordBatch, Table, concat_tables
 import pyarrow.lib as ext
-from .util import _deprecate_nthreads
 
 
 try:
@@ -68,8 +67,7 @@ class FeatherReader(ext.FeatherReader):
         table = Table.from_arrays(columns, names=names)
         return table
 
-    def read_pandas(self, columns=None, nthreads=None, use_threads=False):
-        use_threads = _deprecate_nthreads(use_threads, nthreads)
+    def read_pandas(self, columns=None, use_threads=True):
         return self.read_table(columns=columns).to_pandas(
             use_threads=use_threads)
 
@@ -145,7 +143,7 @@ class FeatherDataset(object):
                              .format(piece, self.schema,
                                      table.schema))
 
-    def read_pandas(self, columns=None, nthreads=None, use_threads=False):
+    def read_pandas(self, columns=None, use_threads=True):
         """
         Read multiple Parquet files as a single pandas DataFrame
 
@@ -153,15 +151,14 @@ class FeatherDataset(object):
         ----------
         columns : List[str]
             Names of columns to read from the file
-        nthreads : int, default 1
-            Number of columns to read in parallel.
+        use_threads : boolean, default True
+            Use multiple threads when converting to pandas
 
         Returns
         -------
         pandas.DataFrame
             Content of the file as a pandas DataFrame (of columns)
         """
-        use_threads = _deprecate_nthreads(use_threads, nthreads)
         return self.read_table(columns=columns).to_pandas(
             use_threads=use_threads)
 
@@ -192,7 +189,7 @@ def write_feather(df, dest):
         raise
 
 
-def read_feather(source, columns=None, nthreads=None, use_threads=False):
+def read_feather(source, columns=None, use_threads=True):
     """
     Read a pandas.DataFrame from Feather format
 
@@ -202,14 +199,13 @@ def read_feather(source, columns=None, nthreads=None, use_threads=False):
     columns : sequence, optional
         Only read a specific set of columns. If not provided, all columns are
         read
-    use_threads: bool, default False
+    use_threads: bool, default True
         Whether to parallelize reading using multiple threads
 
     Returns
     -------
     df : pandas.DataFrame
     """
-    use_threads = _deprecate_nthreads(use_threads, nthreads)
     reader = FeatherReader(source)
     return reader.read_pandas(columns=columns, use_threads=use_threads)
 

python/pyarrow/ipc.py

@@ -24,7 +24,6 @@ from pyarrow.lib import (Message, MessageReader,  # noqa
                          read_tensor, write_tensor,
                          get_record_batch_size, get_tensor_size)
 import pyarrow.lib as lib
-from .util import _deprecate_nthreads
 
 
 class _ReadPandasOption(object):
@@ -169,22 +168,20 @@ def serialize_pandas(df, nthreads=None, preserve_index=True):
     return sink.getvalue()
 
 
-def deserialize_pandas(buf, nthreads=None, use_threads=False):
+def deserialize_pandas(buf, use_threads=True):
     """Deserialize a buffer protocol compatible object into a pandas DataFrame.
 
     Parameters
     ----------
     buf : buffer
         An object compatible with the buffer protocol
-    use_threads: boolean, default False
+    use_threads: boolean, default True
         Whether to parallelize the conversion using multiple threads
 
     Returns
     -------
     df : pandas.DataFrame
     """
-    use_threads = _deprecate_nthreads(use_threads, nthreads)
-
     buffer_reader = pa.BufferReader(buf)
     reader = pa.RecordBatchStreamReader(buffer_reader)
     table = reader.read_all()

python/pyarrow/table.pxi

@@ -27,8 +27,6 @@ except ImportError:
 else:
     import pyarrow.pandas_compat as pdcompat
 
-from .util import _deprecate_nthreads
-
 
 cdef class ChunkedArray:
     """
@@ -865,15 +863,19 @@ cdef class RecordBatch:
             entries.append((name, column))
         return OrderedDict(entries)
 
-    def to_pandas(self, nthreads=None, use_threads=False):
+    def to_pandas(self, use_threads=True):
         """
         Convert the arrow::RecordBatch to a pandas DataFrame
 
+        Parameters
+        ----------
+        use_threads : boolean, default True
+            Use multiple threads for conversion
+
         Returns
         -------
         pandas.DataFrame
         """
-        use_threads = _deprecate_nthreads(use_threads, nthreads)
         return Table.from_batches([self]).to_pandas(use_threads=use_threads)
 
     @classmethod
@@ -1289,9 +1291,9 @@ cdef class Table:
 
         return result
 
-    def to_pandas(self, nthreads=None, strings_to_categorical=False,
+    def to_pandas(self, strings_to_categorical=False,
                   memory_pool=None, zero_copy_only=False, categories=None,
-                  integer_object_nulls=False, use_threads=False):
+                  integer_object_nulls=False, use_threads=True):
         """
         Convert the arrow::Table to a pandas DataFrame
 
@@ -1308,7 +1310,7 @@ cdef class Table:
             List of columns that should be returned as pandas.Categorical
         integer_object_nulls : boolean, default False
             Cast integers with nulls to objects
-        use_threads: boolean, default False
+        use_threads: boolean, default True
             Whether to parallelize the conversion using multiple threads
 
         Returns
@@ -1318,8 +1320,6 @@ cdef class Table:
         cdef:
             PandasOptions options
 
-        use_threads = _deprecate_nthreads(use_threads, nthreads)
-
         options = PandasOptions(
             strings_to_categorical=strings_to_categorical,
             zero_copy_only=zero_copy_only,

python/pyarrow/util.py

@@ -35,13 +35,3 @@ def _deprecate_api(old_name, new_name, api, next_version):
         warnings.warn(msg, FutureWarning)
         return api(*args)
     return wrapper
-
-
-def _deprecate_nthreads(use_threads, nthreads):
-    if nthreads is not None:
-        warnings.warn("`nthreads` argument is deprecated, "
-                      "pass `use_threads` instead", FutureWarning,
-                      stacklevel=3)
-        if nthreads > 1:
-            use_threads = True
-    return use_threads