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

pca_metrics_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 import daal.algorithms.pca as pca
48 import daal.algorithms.pca.quality_metric_set as quality_metric_set
49 from daal.algorithms.pca.quality_metric import explained_variance
50 from daal.data_management import (
51  DataSourceIface, FileDataSource, HomogenNumericTable, MergedNumericTable,
52  NumericTableIface, BlockDescriptor, readWrite
53 )
54 
55 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
56 if utils_folder not in sys.path:
57  sys.path.insert(0, utils_folder)
58 from utils import printNumericTable
59 
60 datasetFileName = os.path.join('..', 'data', 'batch', 'pca_normalized.csv')
61 nVectors = 1000
62 nComponents = 5
63 
64 qmsResult = None
65 eigenData = None
66 
67 def trainModel():
68  global eigenData
69 
70  # Initialize FileDataSource to retrieve the input data from a .csv file
71  dataSource = FileDataSource(
72  datasetFileName,
73  DataSourceIface.doAllocateNumericTable,
74  DataSourceIface.doDictionaryFromContext
75  )
76 
77  # Retrieve the data from the input file
78  dataSource.loadDataBlock(nVectors)
79 
80  # Create an algorithm for principal component analysis using the SVD method
81  algorithm = pca.Batch(method=pca.svdDense)
82 
83  # Set the algorithm input data
84  algorithm.input.setDataset(pca.data, dataSource.getNumericTable())
85 
86  # Compute results of the PCA algorithm
87  result = algorithm.compute()
88  eigenData = result.get(pca.eigenvalues)
89 
90 def testPcaQuality():
91  global qmsResult
92 
93  # Create a quality metric set object to compute quality metrics of the PCA algorithm
94  qualityMetricSet = quality_metric_set.Batch(nComponents)
95  explainedVariances = explained_variance.Input.downCast(qualityMetricSet.getInputDataCollection().getInput(quality_metric_set.explainedVariancesMetrics))
96  explainedVariances.setInput(explained_variance.eigenvalues, eigenData)
97 
98  # Compute quality metrics
99  qualityMetricSet.compute()
100 
101  # Retrieve the quality metrics
102  qmsResult = qualityMetricSet.getResultCollection()
103 
104 def printResults():
105  print ("Quality metrics for PCA")
106  result = explained_variance.Result.downCast(qmsResult.getResult(quality_metric_set.explainedVariancesMetrics))
107  printNumericTable(result.getResult(explained_variance.explainedVariances), "Explained variances:")
108  printNumericTable(result.getResult(explained_variance.explainedVariancesRatios), "Explained variances ratios:")
109  printNumericTable(result.getResult(explained_variance.noiseVariance), "Noise variance:")
110 
111 if __name__ == "__main__":
112  trainModel()
113  testPcaQuality()
114  printResults()

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