Add embedding script

RubixML · Mar 6, 2022 · 5472285 · 5472285
1 parent 3e4cc69
commit 5472285
Show file tree

Hide file tree

Showing 4 changed files with 25 additions and 0 deletions.
diff --git a/.gitignore b/.gitignore
@@ -1,6 +1,7 @@
 /vendor
 composer.lock
 report.json
+embedding.csv
 *.model
 .vscode
 .vs
diff --git a/composer.json b/composer.json
@@ -21,6 +21,7 @@
         "rubix/ml": "^1.0"
     },
     "scripts": {
+        "explore": "@php explore.php",
         "train": "@php train.php",
         "simulate": "@php simulate.php"
     },

diff --git a/docs/images/original-divorce-dataset-2-d-pca-embedding.png b/docs/images/original-divorce-dataset-2-d-pca-embedding.png
diff --git a/explore.php b/explore.php
@@ -0,0 +1,23 @@
+<?php
+
+include __DIR__ . '/vendor/autoload.php';
+
+use Rubix\ML\Loggers\Screen;
+use Rubix\ML\Datasets\Labeled;
+use Rubix\ML\Extractors\NDJSON;
+use Rubix\ML\Transformers\PrincipalComponentAnalysis;
+use Rubix\ML\Extractors\CSV;
+
+ini_set('memory_limit', '-1');
+
+$logger = new Screen();
+
+$logger->info('Loading data into memory');
+
+$dataset = Labeled::fromIterator(new NDJSON('dataset.ndjson'));
+
+$embedder = new PrincipalComponentAnalysis(2);
+
+$dataset->apply($embedder)->exportTo(new CSV('embedding.csv'));
+
+$logger->info('Embedding saved to embedding.csv');