1 |
efrain |
1 |
<?php
|
|
|
2 |
// This file is part of Moodle - http://moodle.org/
|
|
|
3 |
//
|
|
|
4 |
// Moodle is free software: you can redistribute it and/or modify
|
|
|
5 |
// it under the terms of the GNU General Public License as published by
|
|
|
6 |
// the Free Software Foundation, either version 3 of the License, or
|
|
|
7 |
// (at your option) any later version.
|
|
|
8 |
//
|
|
|
9 |
// Moodle is distributed in the hope that it will be useful,
|
|
|
10 |
// but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
11 |
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
12 |
// GNU General Public License for more details.
|
|
|
13 |
//
|
|
|
14 |
// You should have received a copy of the GNU General Public License
|
|
|
15 |
// along with Moodle. If not, see <http://www.gnu.org/licenses/>.
|
|
|
16 |
|
|
|
17 |
namespace core_analytics;
|
|
|
18 |
|
|
|
19 |
/**
|
|
|
20 |
* Unit tests for the dataset manager.
|
|
|
21 |
*
|
|
|
22 |
* @package core_analytics
|
|
|
23 |
* @copyright 2017 David Monllaó {@link http://www.davidmonllao.com}
|
|
|
24 |
* @license http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
|
|
|
25 |
*/
|
|
|
26 |
class dataset_manager_test extends \advanced_testcase {
|
|
|
27 |
|
|
|
28 |
/** @var array Store dataset top rows. */
|
|
|
29 |
protected array $sharedtoprows = [];
|
|
|
30 |
|
|
|
31 |
/**
|
|
|
32 |
* setUp
|
|
|
33 |
*
|
|
|
34 |
* @return null
|
|
|
35 |
*/
|
|
|
36 |
public function setUp(): void {
|
|
|
37 |
$this->resetAfterTest(true);
|
|
|
38 |
|
|
|
39 |
$this->sharedtoprows = array(
|
|
|
40 |
array('var1', 'var2'),
|
|
|
41 |
array('value1', 'value2'),
|
|
|
42 |
array('header1', 'header2')
|
|
|
43 |
);
|
|
|
44 |
}
|
|
|
45 |
|
|
|
46 |
/**
|
|
|
47 |
* test_create_dataset
|
|
|
48 |
*
|
|
|
49 |
* @return null
|
|
|
50 |
*/
|
11 |
efrain |
51 |
public function test_create_dataset(): void {
|
1 |
efrain |
52 |
|
|
|
53 |
$dataset1 = new \core_analytics\dataset_manager(1, 1, 'whatever', \core_analytics\dataset_manager::LABELLED_FILEAREA, false);
|
|
|
54 |
$dataset1data = array_merge($this->sharedtoprows, array(array('yeah', 'yeah', 'yeah')));
|
|
|
55 |
$f1 = $dataset1->store($dataset1data);
|
|
|
56 |
|
|
|
57 |
$f1contents = $f1->get_content();
|
|
|
58 |
$this->assertStringContainsString('yeah', $f1contents);
|
|
|
59 |
$this->assertStringContainsString('var1', $f1contents);
|
|
|
60 |
$this->assertStringContainsString('value1', $f1contents);
|
|
|
61 |
$this->assertStringContainsString('header1', $f1contents);
|
|
|
62 |
}
|
|
|
63 |
|
|
|
64 |
/**
|
|
|
65 |
* test_merge_datasets
|
|
|
66 |
*
|
|
|
67 |
* @return null
|
|
|
68 |
*/
|
11 |
efrain |
69 |
public function test_merge_datasets(): void {
|
1 |
efrain |
70 |
|
|
|
71 |
$dataset1 = new \core_analytics\dataset_manager(1, 1, 'whatever', \core_analytics\dataset_manager::LABELLED_FILEAREA, false);
|
|
|
72 |
$dataset1data = array_merge($this->sharedtoprows, array(array('yeah', 'yeah', 'yeah')));
|
|
|
73 |
$f1 = $dataset1->store($dataset1data);
|
|
|
74 |
|
|
|
75 |
$dataset2 = new \core_analytics\dataset_manager(1, 2, 'whatever', \core_analytics\dataset_manager::LABELLED_FILEAREA, false);
|
|
|
76 |
$dataset2data = array_merge($this->sharedtoprows, array(array('no', 'no', 'no')));
|
|
|
77 |
$f2 = $dataset2->store($dataset2data);
|
|
|
78 |
|
|
|
79 |
$files = array($f1, $f2);
|
|
|
80 |
$merged = \core_analytics\dataset_manager::merge_datasets($files, 1, 'whatever',
|
|
|
81 |
\core_analytics\dataset_manager::LABELLED_FILEAREA);
|
|
|
82 |
|
|
|
83 |
$mergedfilecontents = $merged->get_content();
|
|
|
84 |
$this->assertStringContainsString('yeah', $mergedfilecontents);
|
|
|
85 |
$this->assertStringContainsString('no', $mergedfilecontents);
|
|
|
86 |
$this->assertStringContainsString('var1', $mergedfilecontents);
|
|
|
87 |
$this->assertStringContainsString('value1', $mergedfilecontents);
|
|
|
88 |
$this->assertStringContainsString('header1', $mergedfilecontents);
|
|
|
89 |
}
|
|
|
90 |
|
|
|
91 |
/**
|
|
|
92 |
* test_get_pending_files
|
|
|
93 |
*
|
|
|
94 |
* @return null
|
|
|
95 |
*/
|
11 |
efrain |
96 |
public function test_get_pending_files(): void {
|
1 |
efrain |
97 |
global $DB;
|
|
|
98 |
|
|
|
99 |
$this->resetAfterTest();
|
|
|
100 |
|
|
|
101 |
$fakemodelid = 123;
|
|
|
102 |
$timesplittingids = array(
|
|
|
103 |
'\core\analytics\time_splitting\quarters',
|
|
|
104 |
'\core\analytics\time_splitting\quarters_accum',
|
|
|
105 |
);
|
|
|
106 |
|
|
|
107 |
// No files.
|
|
|
108 |
$this->assertEmpty(\core_analytics\dataset_manager::get_pending_files($fakemodelid, true, $timesplittingids));
|
|
|
109 |
$this->assertEmpty(\core_analytics\dataset_manager::get_pending_files($fakemodelid, false, $timesplittingids));
|
|
|
110 |
|
|
|
111 |
// We will reuse this analysable file to create training and prediction datasets (analysable level files are
|
|
|
112 |
// merged into training and prediction files).
|
|
|
113 |
$analysabledataset = new \core_analytics\dataset_manager($fakemodelid, 1, 'whatever',
|
|
|
114 |
\core_analytics\dataset_manager::LABELLED_FILEAREA, false);
|
|
|
115 |
$analysabledatasetdata = array_merge($this->sharedtoprows, array(array('yeah', 'yeah', 'yeah')));
|
|
|
116 |
$file = $analysabledataset->store($analysabledatasetdata);
|
|
|
117 |
|
|
|
118 |
// Evaluation files ignored.
|
|
|
119 |
$evaluationdataset = \core_analytics\dataset_manager::merge_datasets(array($file), $fakemodelid,
|
|
|
120 |
'\core\analytics\time_splitting\quarters', \core_analytics\dataset_manager::LABELLED_FILEAREA, true);
|
|
|
121 |
|
|
|
122 |
$this->assertEmpty(\core_analytics\dataset_manager::get_pending_files($fakemodelid, true, $timesplittingids));
|
|
|
123 |
$this->assertEmpty(\core_analytics\dataset_manager::get_pending_files($fakemodelid, false, $timesplittingids));
|
|
|
124 |
|
|
|
125 |
// Training and prediction files are not mixed up.
|
|
|
126 |
$trainingfile1 = \core_analytics\dataset_manager::merge_datasets(array($file), $fakemodelid,
|
|
|
127 |
'\core\analytics\time_splitting\quarters', \core_analytics\dataset_manager::LABELLED_FILEAREA, false);
|
|
|
128 |
$this->waitForSecond();
|
|
|
129 |
$trainingfile2 = \core_analytics\dataset_manager::merge_datasets(array($file), $fakemodelid,
|
|
|
130 |
'\core\analytics\time_splitting\quarters', \core_analytics\dataset_manager::LABELLED_FILEAREA, false);
|
|
|
131 |
|
|
|
132 |
$bytimesplitting = \core_analytics\dataset_manager::get_pending_files($fakemodelid, true, $timesplittingids);
|
|
|
133 |
$this->assertFalse(isset($bytimesplitting['\core\analytics\time_splitting\quarters_accum']));
|
|
|
134 |
$this->assertCount(2, $bytimesplitting['\core\analytics\time_splitting\quarters']);
|
|
|
135 |
$this->assertEmpty(\core_analytics\dataset_manager::get_pending_files($fakemodelid, false, $timesplittingids));
|
|
|
136 |
|
|
|
137 |
$predictionfile = \core_analytics\dataset_manager::merge_datasets(array($file), $fakemodelid,
|
|
|
138 |
'\core\analytics\time_splitting\quarters', \core_analytics\dataset_manager::UNLABELLED_FILEAREA, false);
|
|
|
139 |
$bytimesplitting = \core_analytics\dataset_manager::get_pending_files($fakemodelid, false, $timesplittingids);
|
|
|
140 |
$this->assertFalse(isset($bytimesplitting['\core\analytics\time_splitting\quarters_accum']));
|
|
|
141 |
$this->assertCount(1, $bytimesplitting['\core\analytics\time_splitting\quarters']);
|
|
|
142 |
|
|
|
143 |
// Already used for training and prediction are discarded.
|
|
|
144 |
$usedfile = (object)['modelid' => $fakemodelid, 'fileid' => $trainingfile1->get_id(), 'action' => 'trained',
|
|
|
145 |
'time' => time()];
|
|
|
146 |
$DB->insert_record('analytics_used_files', $usedfile);
|
|
|
147 |
$bytimesplitting = \core_analytics\dataset_manager::get_pending_files($fakemodelid, true, $timesplittingids);
|
|
|
148 |
$this->assertCount(1, $bytimesplitting['\core\analytics\time_splitting\quarters']);
|
|
|
149 |
|
|
|
150 |
$usedfile->fileid = $predictionfile->get_id();
|
|
|
151 |
$usedfile->action = 'predicted';
|
|
|
152 |
$DB->insert_record('analytics_used_files', $usedfile);
|
|
|
153 |
$this->assertEmpty(\core_analytics\dataset_manager::get_pending_files($fakemodelid, false, $timesplittingids));
|
|
|
154 |
}
|
|
|
155 |
}
|