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

mn_naive_bayes_csr_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.multinomial_naive_bayes import prediction, training
48 from daal.algorithms import classifier
49 from daal.data_management import FileDataSource, DataSourceIface
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 printNumericTables, createSparseTable
55 
56 DAAL_PREFIX = os.path.join('..', 'data')
57 
58 # Input data set parameters
59 trainDatasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_train_csr.csv')
60 trainGroundTruthFileName = os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_train_labels.csv')
61 
62 testDatasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_test_csr.csv')
63 testGroundTruthFileName = os.path.join(DAAL_PREFIX, 'batch', 'naivebayes_test_labels.csv')
64 
65 nTrainObservations = 8000
66 nTestObservations = 2000
67 nClasses = 20
68 
69 trainingResult = None
70 predictionResult = None
71 
72 
73 def trainModel():
74  global trainingResult
75 
76  # Initialize FileDataSource<CSVFeatureManager> to retrieve the input data from a .csv file
77  trainGroundTruthSource = FileDataSource(
78  trainGroundTruthFileName,
79  DataSourceIface.doAllocateNumericTable,
80  DataSourceIface.doDictionaryFromContext
81  )
82 
83  # Retrieve the data from input files
84  trainData = createSparseTable(trainDatasetFileName)
85  trainGroundTruthSource.loadDataBlock(nTrainObservations)
86 
87  # Create an algorithm object to train the Naive Bayes model
88  algorithm = training.Batch(nClasses, method=training.fastCSR)
89 
90  # Pass a training data set and dependent values to the algorithm
91  algorithm.input.set(classifier.training.data, trainData)
92  algorithm.input.set(classifier.training.labels, trainGroundTruthSource.getNumericTable())
93 
94  # Build the Naive Bayes model and retrieve the algorithm results
95  trainingResult = algorithm.compute()
96 
97 
98 def testModel():
99  global predictionResult
100 
101  # Initialize FileDataSource<CSVFeatureManager> to retrieve the test data from a .csv file
102  testData = createSparseTable(testDatasetFileName)
103 
104  # Create an algorithm object to predict Naive Bayes values
105  algorithm = prediction.Batch(nClasses, method=prediction.fastCSR)
106 
107  # Pass a testing data set and the trained model to the algorithm
108  algorithm.input.setTable(classifier.prediction.data, testData)
109  algorithm.input.setModel(classifier.prediction.model, trainingResult.get(classifier.training.model))
110 
111  # Predict Naive Bayes values and retrieve the algorithm results (Result class from classifier.prediction)
112  predictionResult = algorithm.compute()
113 
114 
115 def printResults():
116 
117  testGroundTruth = FileDataSource(
118  testGroundTruthFileName,
119  DataSourceIface.doAllocateNumericTable,
120  DataSourceIface.doDictionaryFromContext
121  )
122 
123  testGroundTruth.loadDataBlock(nTestObservations)
124 
125  printNumericTables(
126  testGroundTruth.getNumericTable(),
127  predictionResult.get(classifier.prediction.prediction),
128  "Ground truth", "Classification results",
129  "NaiveBayes classification results (first 20 observations):", 20, 15, flt64=False
130  )
131 
132 if __name__ == "__main__":
133 
134  trainModel()
135  testModel()
136  printResults()

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