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

impl_als_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.implicit_als.training.init
48 import daal.algorithms.implicit_als.prediction.ratings
49 from daal.algorithms.implicit_als import training, prediction
50 from daal.data_management import FileDataSource, DataSourceIface
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', 'implicit_als_dense.csv')
61 
62 # Algorithm parameters
63 nFactors = 2
64 
65 dataTable = None
66 initialModel = None
67 trainingResult = None
68 
69 
70 def initializeModel():
71  global dataTable, initialModel
72 
73  # Read trainDatasetFileName from a file and create a numeric table to store the input data
74  dataSource = FileDataSource(
75  trainDatasetFileName, DataSourceIface.doAllocateNumericTable,
76  DataSourceIface.doDictionaryFromContext
77  )
78 
79  # Retrieve the input data
80  dataSource.loadDataBlock()
81 
82  dataTable = dataSource.getNumericTable()
83  # Create an algorithm object to initialize the implicit ALS model with the default method
84  initAlgorithm = training.init.Batch()
85  initAlgorithm.parameter.nFactors = nFactors
86 
87  # Pass a training data set and dependent values to the algorithm
88  initAlgorithm.input.set(training.init.data, dataTable)
89  res = initAlgorithm.compute()
90 
91  # Initialize the implicit ALS model
92  initialModel = res.get(training.init.model)
93 
94 
95 def trainModel():
96  global trainingResult
97 
98  # Create an algorithm object to train the implicit ALS model with the default method
99  algorithm = training.Batch()
100 
101  # Pass a training data set and dependent values to the algorithm
102  algorithm.input.setTable(training.data, dataTable)
103  algorithm.input.setModel(training.inputModel, initialModel)
104 
105  algorithm.parameter.nFactors = nFactors
106 
107  # Build the implicit ALS model and retrieve the algorithm results
108  trainingResult = algorithm.compute()
109 
110 
111 def testModel():
112 
113  # Create an algorithm object to predict recommendations of the implicit ALS model
114  algorithm = prediction.ratings.Batch()
115  algorithm.parameter.nFactors = nFactors
116 
117  algorithm.input.set(prediction.ratings.model, trainingResult.get(training.model))
118 
119  res = algorithm.compute()
120  predictedRatings = res.get(prediction.ratings.prediction)
121 
122  printNumericTable(predictedRatings, "Predicted ratings:")
123 
124 if __name__ == "__main__":
125 
126  initializeModel()
127  trainModel()
128  testModel()

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