Python* API Reference for Intel® Data Analytics Acceleration Library 2019

pca_cor_csr_distr.py

1 # file: pca_cor_csr_distr.py
2 #===============================================================================
3 # Copyright 2014-2018 Intel Corporation.
4 #
5 # This software and the related documents are Intel copyrighted materials, and
6 # your use of them is governed by the express license under which they were
7 # provided to you (License). Unless the License provides otherwise, you may not
8 # use, modify, copy, publish, distribute, disclose or transmit this software or
9 # the related documents without Intel's prior written permission.
10 #
11 # This software and the related documents are provided as is, with no express
12 # or implied warranties, other than those that are expressly stated in the
13 # License.
14 #===============================================================================
15 
16 ## <a name="DAAL-EXAMPLE-PY-PCA_CORRELATION_CSR_DISTRIBUTED"></a>
17 ## \example pca_cor_csr_distr.py
18 
19 import os
20 import sys
21 
22 import numpy as np
23 
24 from daal import step1Local, step2Master
25 from daal.algorithms import covariance
26 from daal.algorithms import pca
27 
28 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
29 if utils_folder not in sys.path:
30  sys.path.insert(0, utils_folder)
31 from utils import printNumericTable, createSparseTable
32 
33 DAAL_PREFIX = os.path.join('..', 'data')
34 
35 # Input data set parameters
36 nBlocks = 4
37 datasetFileNames = [
38  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_1.csv'),
39  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_2.csv'),
40  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_3.csv'),
41  os.path.join(DAAL_PREFIX, 'distributed', 'covcormoments_csr_4.csv')
42 ]
43 
44 if __name__ == "__main__":
45 
46  # Create an algorithm for principal component analysis using the correlation method on the master node
47  masterAlgorithm = pca.Distributed(step2Master,fptype=np.float64)
48 
49  for i in range(nBlocks):
50  dataTable = createSparseTable(datasetFileNames[i])
51 
52  # Create algorithm objects to compute a variance-covariance matrix in the distributed processing mode using the default method
53  localAlgorithm = pca.Distributed(step1Local,fptype=np.float64)
54 
55  # Create an algorithm for principal component analysis using the correlation method on the local node
56  localAlgorithm.parameter.covariance = covariance.Distributed(step1Local, fptype=np.float64, method=covariance.fastCSR)
57 
58  # Set input objects for the algorithm
59  localAlgorithm.input.setDataset(pca.data, dataTable)
60 
61  # Compute partial estimates on local nodes
62  # Set local partial results as input for the master-node algorithm
63  masterAlgorithm.input.add(pca.partialResults, localAlgorithm.compute())
64 
65  # Use covariance algorithm for sparse data inside the PCA algorithm
66  masterAlgorithm.parameter.covariance = covariance.Distributed(step2Master, fptype=np.float64, method=covariance.fastCSR)
67 
68  # Merge and finalize PCA decomposition on the master node
69  masterAlgorithm.compute()
70 
71  result = masterAlgorithm.finalizeCompute()
72 
73  # Print the results
74  printNumericTable(result.get(pca.eigenvalues), "Eigenvalues:")
75  printNumericTable(result.get(pca.eigenvectors), "Eigenvectors:")

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