<!--
{
  "availability" : [
    "iOS: 15.0.0 - 17.0.0",
    "iPadOS: 15.0.0 - 17.0.0",
    "macCatalyst: 15.0.0 - 17.0.0",
    "macOS: 10.14.0 - 14.0.0",
    "tvOS: 16.0.0 - 17.0.0",
    "visionOS: 1.0.0 -"
  ],
  "documentType" : "symbol",
  "framework" : "CreateML",
  "identifier" : "/documentation/CreateML/MLClassifierMetrics/init(classificationError:confusion:precisionRecall:)",
  "metadataVersion" : "0.1.0",
  "role" : "Initializer",
  "symbol" : {
    "kind" : "Initializer",
    "modules" : [
      "Create ML"
    ],
    "preciseIdentifier" : "s:8CreateML19MLClassifierMetricsV19classificationError9confusion15precisionRecallACSd_AA11MLDataTableVAHtcfc"
  },
  "title" : "init(classificationError:confusion:precisionRecall:)"
}
-->

# init(classificationError:confusion:precisionRecall:)

Creates empty classifier metrics.

```
init(classificationError: Double, confusion: MLDataTable, precisionRecall: MLDataTable)
```

## Parameters

`classificationError`

The fraction of incorrectly labeled examples.

`confusion`

A confusion matrix describing the classifications for each category.

`precisionRecall`

A two-dimensional table describing the precision and recall for each category.

## Discussion

You typically don’t initialize metrics directly. Instead you get metrics about your model after training. For
example, when you train an [`MLClassifier`](/documentation/CreateML/MLClassifier), you can look at its [`trainingMetrics`](/documentation/CreateML/MLClassifier/trainingMetrics) and
[`validationMetrics`](/documentation/CreateML/MLClassifier/validationMetrics) properties. Additionally, you can check the performance on a test set with
the [`evaluation(on:)`](/documentation/CreateML/MLClassifier/evaluation(on:)-6433y) method.

> Warning: This initializer should not be used, it creates an empty instance.

---

Copyright &copy; 2026 Apple Inc. All rights reserved. | [Terms of Use](https://www.apple.com/legal/internet-services/terms/site.html) | [Privacy Policy](https://www.apple.com/privacy/privacy-policy)