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

impl_als_csr_batch.py

1 # file: impl_als_csr_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.prediction.ratings as ratings
49 import daal.algorithms.implicit_als.training as training
50 import daal.algorithms.implicit_als.training.init as init
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, createSparseTable
56 
57 DAAL_PREFIX = os.path.join('..', 'data')
58 
59 # Input data set parameters
60 trainDatasetFileName = os.path.join(DAAL_PREFIX, 'batch', 'implicit_als_csr.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 initialModel, dataTable
72 
73  # Read trainDatasetFileName from a file and create a numeric table to store the input data
74  dataTable = createSparseTable(trainDatasetFileName)
75 
76  # Create an algorithm object to initialize the implicit ALS model with the default method
77  initAlgorithm = init.Batch(method=init.fastCSR)
78  initAlgorithm.parameter.nFactors = nFactors
79 
80  # Pass a training data set and dependent values to the algorithm
81  initAlgorithm.input.set(init.data, dataTable)
82 
83  # Initialize the implicit ALS model
84  res = initAlgorithm.compute()
85  # (Result class from implicit_als.training.init)
86  initialModel = res.get(init.model)
87 
88 
89 def trainModel():
90  global trainingResult
91 
92  # Create an algorithm object to train the implicit ALS model with the default method
93  algorithm = training.Batch(method=training.fastCSR)
94 
95  # Pass a training data set and dependent values to the algorithm
96  algorithm.input.setTable(training.data, dataTable)
97  algorithm.input.setModel(training.inputModel, initialModel)
98 
99  algorithm.parameter.nFactors = nFactors
100 
101  # Build the implicit ALS model
102  # Retrieve the algorithm results
103  trainingResult = algorithm.compute()
104 
105 
106 def testModel():
107 
108  # Create an algorithm object to predict recommendations of the implicit ALS model
109  algorithm = ratings.Batch()
110  algorithm.parameter.nFactors = nFactors
111 
112  algorithm.input.set(ratings.model, trainingResult.get(training.model))
113 
114  res = algorithm.compute()
115 
116  predictedRatings = res.get(ratings.prediction)
117 
118  printNumericTable(predictedRatings, "Predicted ratings:")
119 
120 if __name__ == "__main__":
121 
122  initializeModel()
123  trainModel()
124  testModel()

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