Python* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 2

impl_als_dense_batch.py

1 # file: impl_als_dense_batch.py
2 #===============================================================================
3 # Copyright 2014-2018 Intel Corporation
4 # All Rights Reserved.
5 #
6 # If this software was obtained under the Intel Simplified Software License,
7 # the following terms apply:
8 #
9 # The source code, information and material ("Material") contained herein is
10 # owned by Intel Corporation or its suppliers or licensors, and title to such
11 # Material remains with Intel Corporation or its suppliers or licensors. The
12 # Material contains proprietary information of Intel or its suppliers and
13 # licensors. The Material is protected by worldwide copyright laws and treaty
14 # provisions. No part of the Material may be used, copied, reproduced,
15 # modified, published, uploaded, posted, transmitted, distributed or disclosed
16 # in any way without Intel's prior express written permission. No license under
17 # any patent, copyright or other intellectual property rights in the Material
18 # is granted to or conferred upon you, either expressly, by implication,
19 # inducement, estoppel or otherwise. Any license under such intellectual
20 # property rights must be express and approved by Intel in writing.
21 #
22 # Unless otherwise agreed by Intel in writing, you may not remove or alter this
23 # notice or any other notice embedded in Materials by Intel or Intel's
24 # suppliers or licensors in any way.
25 #
26 #
27 # If this software was obtained under the Apache License, Version 2.0 (the
28 # "License"), the following terms apply:
29 #
30 # You may not use this file except in compliance with the License. You may
31 # obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
32 #
33 #
34 # Unless required by applicable law or agreed to in writing, software
35 # distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
36 # WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
37 #
38 # See the License for the specific language governing permissions and
39 # limitations under the License.
40 #===============================================================================
41 
42 
43 
44 
45 import os
46 import sys
47 
48 import daal.algorithms.implicit_als.training.init
49 import daal.algorithms.implicit_als.prediction.ratings
50 from daal.algorithms.implicit_als import training, prediction
51 from daal.data_management import FileDataSource, DataSourceIface
52 
53 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
54 if utils_folder not in sys.path:
55  sys.path.insert(0, utils_folder)
56 from utils import printNumericTable
57 
58 DAAL_PREFIX = os.path.join('..', 'data')
59 
60 # Input data set parameters
61 trainDatasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'implicit_als_dense.csv')
62 
63 # Algorithm parameters
64 nFactors = 2
65 
66 dataTable = None
67 initialModel = None
68 trainingResult = None
69 
70 
71 def initializeModel():
72  global dataTable, initialModel
73 
74  # Read trainDatasetFileName from a file and create a numeric table to store the input data
75  dataSource = FileDataSource(
76  trainDatasetFileName, DataSourceIface.doAllocateNumericTable,
77  DataSourceIface.doDictionaryFromContext
78  )
79 
80  # Retrieve the input data
81  dataSource.loadDataBlock()
82 
83  dataTable = dataSource.getNumericTable()
84  # Create an algorithm object to initialize the implicit ALS model with the default method
85  initAlgorithm = training.init.Batch()
86  initAlgorithm.parameter.nFactors = nFactors
87 
88  # Pass a training data set and dependent values to the algorithm
89  initAlgorithm.input.set(training.init.data, dataTable)
90  res = initAlgorithm.compute()
91 
92  # Initialize the implicit ALS model
93  initialModel = res.get(training.init.model)
94 
95 
96 def trainModel():
97  global trainingResult
98 
99  # Create an algorithm object to train the implicit ALS model with the default method
100  algorithm = training.Batch()
101 
102  # Pass a training data set and dependent values to the algorithm
103  algorithm.input.setTable(training.data, dataTable)
104  algorithm.input.setModel(training.inputModel, initialModel)
105 
106  algorithm.parameter.nFactors = nFactors
107 
108  # Build the implicit ALS model and retrieve the algorithm results
109  trainingResult = algorithm.compute()
110 
111 
112 def testModel():
113 
114  # Create an algorithm object to predict recommendations of the implicit ALS model
115  algorithm = prediction.ratings.Batch()
116  algorithm.parameter.nFactors = nFactors
117 
118  algorithm.input.set(prediction.ratings.model, trainingResult.get(training.model))
119 
120  res = algorithm.compute()
121  predictedRatings = res.get(prediction.ratings.prediction)
122 
123  printNumericTable(predictedRatings, "Predicted ratings:")
124 
125 if __name__ == "__main__":
126 
127  initializeModel()
128  trainModel()
129  testModel()

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