forked from apache/arrow
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathserialization.py
More file actions
146 lines (111 loc) · 5.16 KB
/
serialization.py
File metadata and controls
146 lines (111 loc) · 5.16 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
from collections import OrderedDict, defaultdict
import sys
import numpy as np
from pyarrow import serialize_pandas, deserialize_pandas
from pyarrow.lib import _default_serialization_context
def register_default_serialization_handlers(serialization_context):
# ----------------------------------------------------------------------
# Set up serialization for primitive datatypes
# TODO(pcm): This is currently a workaround until arrow supports
# arbitrary precision integers. This is only called on long integers,
# see the associated case in the append method in python_to_arrow.cc
serialization_context.register_type(
int, "int",
custom_serializer=lambda obj: str(obj),
custom_deserializer=lambda data: int(data))
if (sys.version_info < (3, 0)):
serialization_context.register_type(
long, "long", # noqa: F821
custom_serializer=lambda obj: str(obj),
custom_deserializer=lambda data: long(data)) # noqa: F821
def _serialize_ordered_dict(obj):
return list(obj.keys()), list(obj.values())
def _deserialize_ordered_dict(data):
return OrderedDict(zip(data[0], data[1]))
serialization_context.register_type(
OrderedDict, "OrderedDict",
custom_serializer=_serialize_ordered_dict,
custom_deserializer=_deserialize_ordered_dict)
def _serialize_default_dict(obj):
return list(obj.keys()), list(obj.values()), obj.default_factory
def _deserialize_default_dict(data):
return defaultdict(data[2], zip(data[0], data[1]))
serialization_context.register_type(
defaultdict, "defaultdict",
custom_serializer=_serialize_default_dict,
custom_deserializer=_deserialize_default_dict)
serialization_context.register_type(
type(lambda: 0), "function",
pickle=True)
serialization_context.register_type(type, "type", pickle=True)
# ----------------------------------------------------------------------
# Set up serialization for numpy with dtype object (primitive types are
# handled efficiently with Arrow's Tensor facilities, see
# python_to_arrow.cc)
def _serialize_numpy_array(obj):
return obj.tolist(), obj.dtype.str
def _deserialize_numpy_array(data):
return np.array(data[0], dtype=np.dtype(data[1]))
serialization_context.register_type(
np.ndarray, 'np.array',
custom_serializer=_serialize_numpy_array,
custom_deserializer=_deserialize_numpy_array)
# ----------------------------------------------------------------------
# Set up serialization for pandas Series and DataFrame
try:
import pandas as pd
def _serialize_pandas_series(obj):
return serialize_pandas(pd.DataFrame({obj.name: obj}))
def _deserialize_pandas_series(data):
deserialized = deserialize_pandas(data)
return deserialized[deserialized.columns[0]]
def _serialize_pandas_dataframe(obj):
return serialize_pandas(obj)
def _deserialize_pandas_dataframe(data):
return deserialize_pandas(data)
serialization_context.register_type(
pd.Series, 'pd.Series',
custom_serializer=_serialize_pandas_series,
custom_deserializer=_deserialize_pandas_series)
serialization_context.register_type(
pd.DataFrame, 'pd.DataFrame',
custom_serializer=_serialize_pandas_dataframe,
custom_deserializer=_deserialize_pandas_dataframe)
except ImportError:
# no pandas
pass
# ----------------------------------------------------------------------
# Set up serialization for pytorch tensors
try:
import torch
def _serialize_torch_tensor(obj):
return obj.numpy()
def _deserialize_torch_tensor(data):
return torch.from_numpy(data)
for t in [torch.FloatTensor, torch.DoubleTensor, torch.HalfTensor,
torch.ByteTensor, torch.CharTensor, torch.ShortTensor,
torch.IntTensor, torch.LongTensor]:
serialization_context.register_type(
t, "torch." + t.__name__,
custom_serializer=_serialize_torch_tensor,
custom_deserializer=_deserialize_torch_tensor)
except ImportError:
# no torch
pass
register_default_serialization_handlers(_default_serialization_context)