Python* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 3

mn_naive_bayes_dense_distr.py

1 # file: mn_naive_bayes_dense_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 
17 
18 
19 import os
20 import sys
21 
22 from daal import step1Local, step2Master
23 from daal.algorithms.multinomial_naive_bayes import prediction, training
24 from daal.algorithms import classifier
25 from daal.data_management import (
26  FileDataSource, DataSourceIface, NumericTableIface, HomogenNumericTable, MergedNumericTable
27 )
28 
29 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
30 if utils_folder not in sys.path:
31  sys.path.insert(0, utils_folder)
32 from utils import printNumericTables
33 
34 DAAL_PREFIX = os.path.join('..', 'data')
35 
36 # Input data set parameters
37 trainDatasetFileNames = [
38  os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_train_dense.csv'),
39  os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_train_dense.csv'),
40  os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_train_dense.csv'),
41  os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_train_dense.csv')
42 ]
43 
44 testDatasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_test_dense.csv')
45 
46 nFeatures = 20
47 nClasses = 20
48 nBlocks = 4
49 
50 trainingResult = None
51 predictionResult = None
52 testGroundTruth = None
53 
54 
55 def trainModel():
56  global trainingResult
57 
58  masterAlgorithm = training.Distributed(step2Master, nClasses)
59 
60  for i in range(nBlocks):
61  # Initialize FileDataSource<CSVFeatureManager> to retrieve the input data from a .csv file
62  trainDataSource = FileDataSource(
63  trainDatasetFileNames[i], DataSourceIface.notAllocateNumericTable,
64  DataSourceIface.doDictionaryFromContext
65  )
66  # Create Numeric Tables for training data and labels
67  trainData = HomogenNumericTable(nFeatures, 0, NumericTableIface.doNotAllocate)
68  trainGroundTruth = HomogenNumericTable(1, 0, NumericTableIface.doNotAllocate)
69  mergedData = MergedNumericTable(trainData, trainGroundTruth)
70 
71  # Retrieve the data from the input file
72  trainDataSource.loadDataBlock(mergedData)
73 
74  # Create an algorithm object to train the Naive Bayes model on the local-node data
75  localAlgorithm = training.Distributed(step1Local, nClasses)
76 
77  # Pass a training data set and dependent values to the algorithm
78  localAlgorithm.input.set(classifier.training.data, trainData)
79  localAlgorithm.input.set(classifier.training.labels, trainGroundTruth)
80 
81  # Build the Naive Bayes model on the local node and
82  # Set the local Naive Bayes model as input for the master-node algorithm
83  masterAlgorithm.input.add(training.partialModels, localAlgorithm.compute())
84 
85  # Merge and finalize the Naive Bayes model on the master node
86  masterAlgorithm.compute()
87  trainingResult = masterAlgorithm.finalizeCompute() # Retrieve the algorithm results
88 
89 
90 def testModel():
91  global predictionResult, testGroundTruth
92 
93  # Initialize FileDataSource<CSVFeatureManager> to retrieve the test data from a .csv file
94  testDataSource = FileDataSource(
95  testDatasetFileName, DataSourceIface.notAllocateNumericTable,
96  DataSourceIface.doDictionaryFromContext
97  )
98 
99  # Create Numeric Tables for testing data and labels
100  testData = HomogenNumericTable(nFeatures, 0, NumericTableIface.doNotAllocate)
101  testGroundTruth = HomogenNumericTable(1, 0, NumericTableIface.doNotAllocate)
102  mergedData = MergedNumericTable(testData, testGroundTruth)
103 
104  # Retrieve the data from input file
105  testDataSource.loadDataBlock(mergedData)
106 
107  # Create an algorithm object to predict Naive Bayes values
108  algorithm = prediction.Batch(nClasses)
109 
110  # Pass a testing data set and the trained model to the algorithm
111  algorithm.input.setTable(classifier.prediction.data, testData)
112  algorithm.input.setModel(classifier.prediction.model, trainingResult.get(classifier.training.model))
113 
114  # Predict Naive Bayes values (Result class from classifier.prediction)
115  predictionResult = algorithm.compute() # Retrieve the algorithm results
116 
117 
118 def printResults():
119  printNumericTables(
120  testGroundTruth, predictionResult.get(classifier.prediction.prediction),
121  "Ground truth", "Classification results",
122  "NaiveBayes classification results (first 20 observations):", 20, flt64=False
123  )
124 
125 if __name__ == "__main__":
126 
127  trainModel()
128  testModel()
129  printResults()

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