Added composer installation instructions

.gitignore

@@ -1,5 +1,6 @@
 /vendor
 composer.lock
 report.json
+*.model
 .vscode
 .vs

README.md

@@ -2,21 +2,16 @@
 Use the [K Nearest Neighbors](https://docs.rubixml.com/en/latest/classifiers/k-nearest-neighbors.html) algorithm to predict who of your friends will stay married or get a divorce based on their answers to a 54 question survey about their partner. Included in this project is a 171 sample human-annotated dataset that we'll use to train the learner.
 
 - **Difficulty**: Easy
-- **Training time**: Seconds
-
-## Requirements
-- [PHP](https://php.net) 7.2 or above
+- **Training time**: Less than a minute
 
 ## Installation
-Clone the repository locally using [Git](https://git-scm.com/):
+Clone the project locally using [Composer](https://getcomposer.org/):
 ```sh
-$ git clone https://github.com/RubixML/Divorce
+$ composer create-project rubix/divorce
 ```
 
-Install dependencies using [Composer](https://getcomposer.org/):
-```sh
-$ composer install
-```
+## Requirements
+- [PHP](https://php.net) 7.2 or above
 
 ## Tutorial
 

composer.json

@@ -21,8 +21,13 @@
         "php": ">=7.2",
         "rubix/ml": "^0.1.0-rc2"
     },
+    "scripts": {
+        "train": "@php train.php"
+    },
     "config": {
         "preferred-install": "dist",
         "sort-packages": true
-    }
+    },
+    "minimum-stability": "dev",
+    "prefer-stable": true
 }

train.php

@@ -21,7 +21,11 @@
 
 echo 'Making predictions ...' . PHP_EOL;
 
-$predictions = $estimator->predict($testing);
+$predictions = $estimator->predict($testing->randomize());
+
+echo 'Example predictions:' . PHP_EOL;
+
+print_r(array_slice($predictions, 0, 3));
 
 $metric = new Accuracy();