Python* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 1

em_gmm_dense_batch.py

1 #===============================================================================
2 # Copyright 2014-2017 Intel Corporation
3 # All Rights Reserved.
4 #
5 # If this software was obtained under the Intel Simplified Software License,
6 # the following terms apply:
7 #
8 # The source code, information and material ("Material") contained herein is
9 # owned by Intel Corporation or its suppliers or licensors, and title to such
10 # Material remains with Intel Corporation or its suppliers or licensors. The
11 # Material contains proprietary information of Intel or its suppliers and
12 # licensors. The Material is protected by worldwide copyright laws and treaty
13 # provisions. No part of the Material may be used, copied, reproduced,
14 # modified, published, uploaded, posted, transmitted, distributed or disclosed
15 # in any way without Intel's prior express written permission. No license under
16 # any patent, copyright or other intellectual property rights in the Material
17 # is granted to or conferred upon you, either expressly, by implication,
18 # inducement, estoppel or otherwise. Any license under such intellectual
19 # property rights must be express and approved by Intel in writing.
20 #
21 # Unless otherwise agreed by Intel in writing, you may not remove or alter this
22 # notice or any other notice embedded in Materials by Intel or Intel's
23 # suppliers or licensors in any way.
24 #
25 #
26 # If this software was obtained under the Apache License, Version 2.0 (the
27 # "License"), the following terms apply:
28 #
29 # You may not use this file except in compliance with the License. You may
30 # obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
31 #
32 #
33 # Unless required by applicable law or agreed to in writing, software
34 # distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
35 # WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
36 #
37 # See the License for the specific language governing permissions and
38 # limitations under the License.
39 #===============================================================================
40 
41 
42 
43 
44 import os
45 import sys
46 
47 from daal.algorithms import em_gmm
48 from daal.data_management import FileDataSource, DataSourceIface
49 
50 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
51 if utils_folder not in sys.path:
52  sys.path.insert(0, utils_folder)
53 from utils import printNumericTable
54 
55 DAAL_PREFIX = os.path.join('..', 'data')
56 
57 # Input data set parameters
58 datasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'em_gmm.csv')
59 nComponents = 2
60 
61 if __name__ == "__main__":
62 
63  # Initialize FileDataSource<CSVFeatureManager> to retrieve the input data from a .csv file
64  dataSource = FileDataSource(
65  datasetFileName,
66  DataSourceIface.doAllocateNumericTable,
67  DataSourceIface.doDictionaryFromContext
68  )
69  nFeatures = dataSource.getNumberOfColumns()
70 
71  # Retrieve the data from the input file
72  dataSource.loadDataBlock()
73 
74  # Create algorithm objects to initialize the EM algorithm for the GMM
75  # computing the number of components using the default method
76  initAlgorithm = em_gmm.init.Batch(nComponents)
77 
78  # Set an input data table for the initialization algorithm
79  initAlgorithm.input.set(em_gmm.init.data, dataSource.getNumericTable())
80 
81  # Compute initial values for the EM algorithm for the GMM with the default parameters
82  resultInit = initAlgorithm.compute()
83 
84  # Create algorithm objects for the EM algorithm for the GMM computing the number of components using the default method
85  algorithm = em_gmm.Batch(nComponents)
86 
87  # Set an input data table for the algorithm
88  algorithm.input.setTable(em_gmm.data, dataSource.getNumericTable())
89  algorithm.input.setValues(em_gmm.inputValues, resultInit)
90 
91  # Compute the results of the EM algorithm for the GMM with the default parameters
92  result = algorithm.compute()
93 
94  # Print the results
95  printNumericTable(result.getResult(em_gmm.weights), "Weights")
96  printNumericTable(result.getResult(em_gmm.means), "Means")
97  for i in range(nComponents):
98  printNumericTable(result.getCovariances(em_gmm.covariances, i), "Covariance")

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