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

pca_cor_csr_distr.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 import step1Local, step2Master
48 from daal.algorithms import covariance
49 from daal.algorithms import pca
50 
51 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
52 if utils_folder not in sys.path:
53  sys.path.insert(0, utils_folder)
54 from utils import printNumericTable, createSparseTable
55 
56 DAAL_PREFIX = os.path.join('..', 'data')
57 
58 # Input data set parameters
59 nBlocks = 4
60 datasetFileNames = [
61  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_1.csv'),
62  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_2.csv'),
63  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_3.csv'),
64  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_4.csv')
65 ]
66 
67 if __name__ == "__main__":
68 
69  # Create an algorithm for principal component analysis using the correlation method on the master node
70  masterAlgorithm = pca.Distributed(step2Master)
71 
72  for i in range(nBlocks):
73  dataTable = createSparseTable(datasetFileNames[i])
74 
75  # Create algorithm objects to compute a variance-covariance matrix in the distributed processing mode using the default method
76  localAlgorithm = pca.Distributed(step1Local)
77 
78  # Create an algorithm for principal component analysis using the correlation method on the local node
79  localAlgorithm.parameter.covariance = covariance.Distributed(step1Local, method=covariance.fastCSR)
80 
81  # Set input objects for the algorithm
82  localAlgorithm.input.setDataset(pca.data, dataTable)
83 
84  # Compute partial estimates on local nodes
85  # Set local partial results as input for the master-node algorithm
86  masterAlgorithm.input.add(pca.partialResults, localAlgorithm.compute())
87 
88  # Use covariance algorithm for sparse data inside the PCA algorithm
89  masterAlgorithm.parameter.covariance = covariance.Distributed(step2Master, method=covariance.fastCSR)
90 
91  # Merge and finalize PCA decomposition on the master node
92  masterAlgorithm.compute()
93 
94  result = masterAlgorithm.finalizeCompute()
95 
96  # Print the results
97  printNumericTable(result.get(pca.eigenvalues), "Eigenvalues:")
98  printNumericTable(result.get(pca.eigenvectors), "Eigenvectors:")

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