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

lin_reg_qr_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 from daal.algorithms.linear_regression import training, prediction
48 from daal.data_management import (
49  DataSourceIface, FileDataSource, HomogenNumericTable, MergedNumericTable, NumericTableIface
50 )
51 
52 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
53 if utils_folder not in sys.path:
54  sys.path.insert(0, utils_folder)
55 from utils import printNumericTable
56 
57 DAAL_PREFIX = os.path.join('..', 'data')
58 
59 # Input data set parameters
60 trainDatasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'linear_regression_train.csv')
61 testDatasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'linear_regression_test.csv')
62 
63 nFeatures = 10 # Number of features in training and testing data sets
64 nDependentVariables = 2 # Number of dependent variables that correspond to each observation
65 
66 trainingResult = None
67 predictionResult = None
68 
69 
70 def trainModel():
71  global trainingResult
72 
73  # Initialize FileDataSource<CSVFeatureManager> to retrieve the input data from a .csv file
74  trainDataSource = FileDataSource(
75  trainDatasetFileName, DataSourceIface.notAllocateNumericTable,
76  DataSourceIface.doDictionaryFromContext
77  )
78 
79  # Create Numeric Tables for training data and dependent variables
80  trainData = HomogenNumericTable(nFeatures, 0, NumericTableIface.doNotAllocate)
81  trainDependentVariables = HomogenNumericTable(
82  nDependentVariables, 0, NumericTableIface.doNotAllocate
83  )
84  mergedData = MergedNumericTable(trainData, trainDependentVariables)
85 
86  # Retrieve the data from input file
87  trainDataSource.loadDataBlock(mergedData)
88 
89  # Create an algorithm object to train the multiple linear regression model with a QR decomposition-based method
90  algorithm = training.Batch(method=training.qrDense)
91 
92  # Pass a training data set and dependent values to the algorithm
93  algorithm.input.set(training.data, trainData)
94  algorithm.input.set(training.dependentVariables, trainDependentVariables)
95 
96  # Build the multiple linear regression model and retrieve the algorithm results
97  trainingResult = algorithm.compute()
98  printNumericTable(trainingResult.get(training.model).getBeta(), "Linear Regression coefficients:")
99 
100 
101 def testModel():
102  global predictionResult
103 
104  # Initialize FileDataSource<CSVFeatureManager> to retrieve the test data from a .csv file
105  testDataSource = FileDataSource(
106  testDatasetFileName, DataSourceIface.doAllocateNumericTable,
107  DataSourceIface.doDictionaryFromContext
108  )
109 
110  # Create Numeric Tables for testing data and ground truth values
111  testData = HomogenNumericTable(nFeatures, 0, NumericTableIface.doNotAllocate)
112  testGroundTruth = HomogenNumericTable(nDependentVariables, 0, NumericTableIface.doNotAllocate)
113  mergedData = MergedNumericTable(testData, testGroundTruth)
114 
115  testDataSource.loadDataBlock(mergedData)
116 
117  # Create an algorithm object to predict values of multiple linear regression
118  algorithm = prediction.Batch()
119 
120  # Pass a testing data set and the trained model to the algorithm
121  algorithm.input.setTable(prediction.data, testData)
122  algorithm.input.setModel(prediction.model, trainingResult.get(training.model))
123 
124  # Predict values of multiple linear regression and retrieve the algorithm results
125  predictionResult = algorithm.compute()
126  printNumericTable(predictionResult.get(prediction.prediction), "Linear Regression prediction results: (first 10 rows):", 10)
127  printNumericTable(testGroundTruth, "Ground truth (first 10 rows):", 10)
128 
129 if __name__ == "__main__":
130 
131  trainModel()
132  testModel()

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