Python* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 2

serialization.py

1 # file: serialization.py
2 #===============================================================================
3 # Copyright 2014-2018 Intel Corporation
4 # All Rights Reserved.
5 #
6 # If this software was obtained under the Intel Simplified Software License,
7 # the following terms apply:
8 #
9 # The source code, information and material ("Material") contained herein is
10 # owned by Intel Corporation or its suppliers or licensors, and title to such
11 # Material remains with Intel Corporation or its suppliers or licensors. The
12 # Material contains proprietary information of Intel or its suppliers and
13 # licensors. The Material is protected by worldwide copyright laws and treaty
14 # provisions. No part of the Material may be used, copied, reproduced,
15 # modified, published, uploaded, posted, transmitted, distributed or disclosed
16 # in any way without Intel's prior express written permission. No license under
17 # any patent, copyright or other intellectual property rights in the Material
18 # is granted to or conferred upon you, either expressly, by implication,
19 # inducement, estoppel or otherwise. Any license under such intellectual
20 # property rights must be express and approved by Intel in writing.
21 #
22 # Unless otherwise agreed by Intel in writing, you may not remove or alter this
23 # notice or any other notice embedded in Materials by Intel or Intel's
24 # suppliers or licensors in any way.
25 #
26 #
27 # If this software was obtained under the Apache License, Version 2.0 (the
28 # "License"), the following terms apply:
29 #
30 # You may not use this file except in compliance with the License. You may
31 # obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
32 #
33 #
34 # Unless required by applicable law or agreed to in writing, software
35 # distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
36 # WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
37 #
38 # See the License for the specific language governing permissions and
39 # limitations under the License.
40 #===============================================================================
41 
42 
43 
44 
45 import os
46 import sys
47 
48 import numpy as np
49 
50 from daal.data_management import HomogenNumericTable, FileDataSource, DataSource, InputDataArchive, OutputDataArchive
51 
52 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
53 if utils_folder not in sys.path:
54  sys.path.insert(0, utils_folder)
55 from utils import printNumericTable
56 
57 # Input data set parameters
58 datasetFileName = os.path.join('..', 'data', 'batch', 'serialization.csv')
59 
60 
61 def serializeNumericTable(dataTable):
62 
63  # Create a data archive to serialize the numeric table
64  dataArch = InputDataArchive()
65 
66  # Serialize the numeric table into the data archive
67  dataTable.serialize(dataArch)
68 
69  # Get the length of the serialized data in bytes
70  length = dataArch.getSizeOfArchive()
71 
72  # Store the serialized data in an array
73  buffer = np.zeros(length, dtype=np.ubyte)
74  dataArch.copyArchiveToArray(buffer)
75 
76  return buffer
77 
78 
79 def deserializeNumericTable(buffer):
80 
81  # Create a data archive to deserialize the numeric table
82  dataArch = OutputDataArchive(buffer)
83 
84  # Create a numeric table object
85  dataTable = HomogenNumericTable()
86 
87  # Deserialize the numeric table from the data archive
88  dataTable.deserialize(dataArch)
89 
90  return dataTable
91 
92 
93 if __name__ == "__main__":
94 
95  # Initialize FileDataSource_CSVFeatureManager to retrieve the input data from a .csv file
96  dataSource = FileDataSource(
97  datasetFileName, DataSource.doAllocateNumericTable, DataSource.doDictionaryFromContext
98  )
99 
100  # Retrieve the data from the input file
101  dataSource.loadDataBlock()
102 
103  # Retrieve a numeric table
104  dataTable = dataSource.getNumericTable()
105 
106  # Print the original data
107  printNumericTable(dataTable, "Data before serialization:")
108 
109  # Serialize the numeric table into the memory buffer
110  buffer = serializeNumericTable(dataTable)
111 
112  # Deserialize the numeric table from the memory buffer
113  restoredDataTable = deserializeNumericTable(buffer)
114 
115  # Print the restored data
116  printNumericTable(restoredDataTable, "Data after deserialization:")

For more complete information about compiler optimizations, see our Optimization Notice.