<!--
{
  "availability" : [
    "iOS: -",
    "iPadOS: -",
    "macCatalyst: -",
    "macOS: -",
    "tvOS: -",
    "visionOS: -",
    "watchOS: -"
  ],
  "documentType" : "symbol",
  "framework" : "Accelerate",
  "identifier" : "/documentation/Accelerate/SparseOpaqueSymbolicFactorization",
  "metadataVersion" : "0.1.0",
  "role" : "Structure",
  "symbol" : {
    "kind" : "Structure",
    "modules" : [
      "Accelerate"
    ],
    "preciseIdentifier" : "c:@SA@SparseOpaqueSymbolicFactorization"
  },
  "title" : "SparseOpaqueSymbolicFactorization"
}
-->

# SparseOpaqueSymbolicFactorization

A semi-opaque type that represents symbolic matrix factorization.

```
struct SparseOpaqueSymbolicFactorization
```

## Overview

Represents a symbolic matrix factorization (that is, the pattern of the factors without the values). A single symbolic factorization may be the basis for multiple numerical factorizations of matrices with the same pattern but different nonzero values.

Use the [`SparseCleanup(_:)`](/documentation/Accelerate/SparseCleanup(_:)-6jpd8) function to free resources that these objects hold. The system reference-counts the internal factorize pointer, so it’s safe to destroy this object even if numeric factorizations exist that still depend on it.

## Topics

### Creating an Opaque Symbolic Factorization

[`init()`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/init())

[`init(status:rowCount:columnCount:attributes:blockSize:type:factorization:workspaceSize_Float:workspaceSize_Double:factorSize_Float:factorSize_Double:)`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/init(status:rowCount:columnCount:attributes:blockSize:type:factorization:workspaceSize_Float:workspaceSize_Double:factorSize_Float:factorSize_Double:))

Creates an opaque symbolic factorization.

### Instance Properties

[`status`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/status)

The status of the factorization.

[`SparseStatus_t`](/documentation/Accelerate/SparseStatus_t)

Constants that describe the status of a factorization.

[`type`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/type)

The factorization type.

[`factorization`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/factorization)

A pointer to a private internal representation of the symbolic factor.

[`SparseFactorization_t`](/documentation/Accelerate/SparseFactorization_t)

Constants that define the factorization type.

### Inspecting a Factorization’s Structure

[`attributes`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/attributes)

The attributes of the factorization.

[`SparseAttributes_t`](/documentation/Accelerate/SparseAttributes_t)

A structure that represents the attributes of a matrix.

[`blockSize`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/blockSize)

The block size.

[`columnCount`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/columnCount)

The number of columns.

[`rowCount`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/rowCount)

The number of rows.

[`workspaceSize_Double`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/workspaceSize_Double)

Size, in bytes, of workspace required to perform numerical factorization in doubles.

[`workspaceSize_Float`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/workspaceSize_Float)

Size, in bytes, of workspace required to perform numerical factorization in floats.

[`factorSize_Double`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/factorSize_Double)

Minimum size, in bytes, required to store numerical factors in doubles.

[`factorSize_Float`](/documentation/Accelerate/SparseOpaqueSymbolicFactorization/factorSize_Float)

Minimum size, in bytes, required to store numerical factors in float.


---

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)