# Classifications API

> Given a query and a set of labeled examples, the model will predict the most likely label for the query. Useful as a drop-in replacement for any ML classification or text-to-label task.
> 
> https://beta.openai.com/docs/api-reference/classifications

Classifications is where things get quite interesting. You give it a set of labelled texts, and it can label those texts for you. So lets start with some data, and see 2 ways we could work with that data.

```php
use SiteOrigin\OpenAI\Client;
use SiteOrigin\OpenAI\Engines;

$examples = [
    ["A happy moment", "Positive"],
    ["I am sad.", "Negative"],
    ["I am feeling awesome", "Positive"],
];

$client = new Client($_ENV['OPENAI_API_KEY']);
$result = $client->classifications(Engines::CURIE)->create('It is a raining day :(', $examples);
echo 'This sentence is ' . $result->label;
```

Where the output of `$result` is:

```php
(object) array(
   'completion' => 'cmpl-3J0IkUnjHCqNfBKyindWidBmFvuOI',
   'label' => 'Negative',
   'model' => 'curie:2020-05-03',
   'object' => 'classification',
   'search_model' => 'ada',
   'selected_examples' => 
  array (
    0 => 
    (object) array(
       'document' => 1,
       'label' => 'Negative',
       'text' => 'I am sad.',
    ),
    1 => 
    (object) array(
       'document' => 0,
       'label' => 'Positive',
       'text' => 'A happy moment',
    ),
    2 => 
    (object) array(
       'document' => 2,
       'label' => 'Positive',
       'text' => 'I am feeling awesome',
    ),
  ),
);
```

Of course, you can also upload a file and use that. In production, you'd obviously only upload the file once, then reference it for all future classifications.

```php
use SiteOrigin\OpenAI\Client;
use SiteOrigin\OpenAI\Engines;use SiteOrigin\OpenAI\Files;

$examples = [
    ["text" => "A happy moment", "label" => "Positive"],
    ["text" => "I am sad.", "label" => "Negative"],
    ["text" => "I am feeling awesome", "label" => "Positive"],
];

$client = new Client($_ENV['OPENAI_API_KEY']);
$file = $client->files()->create('data.json', $examples, Files::PURPOSE_CLASSIFICATIONS);

// Wait for the file to process
$result = $client->classifications(Engines::CURIE)->create('It is a raining day :(', $file->id);
```