NetFabric.Numerics.Tensors

Dealing with SIMD in .NET for optimized code can be complex, but this library offers a practical solution. It provides a reusable and highly-optimized iterations on Span<T>, enabling the application of both pre-defined and custom operations to each element.

Using generics, the library accommodates any type embracing generic math.

Within the library, you'll find pre-defined operations such as Sqrt(), Sin(), Negate(), Add(), Divide(), Multiply(), AddMultiply(), Sum(), Average(), and many more.

For custom operations, the library allows the definition of operators through interfaces like IUnaryOperator<T>, IBinaryOperator<T>, ITernaryOperator<T>, or IAggregationOperator<T>. These operators can be applied seamlessly using the Apply() or Aggregate() methods.

Documentation for this library is available at NetFabric.Numerics.Tensors Documentation.

Usage

To use the NetFabric.Numerics.Tensors library:

1. Install the library via NuGet: dotnet add package NetFabric.Numerics.Tensors
2. Import the library in your code: using NetFabric.Numerics.Tensors;
3. Utilize the library's functions for mathematical operations on tensors represented as spans.

Note: Ensure you're on .NET 8 or a later version for compatibility with the NetFabric.Numerics.Tensors library.

The library includes methods tailored for operations involving one, two, or three ReadOnlySpan<T>. Results are provided in a Span<T>, with the condition that the destination Span<T> must be of the same size or larger than the sources. Inplace operations are supported when the destination parameter matches any of the sources.

For example, given a variable data of type Span<int>, the following code snippet replaces each element of the span with its square root:

Tensor.Sqrt(data, data);

Note that since data serves as both the source and destination, the operation is performed in-place.

For variables x, y, and result, all of type Span<float> and of the same size, the following example updates each element in result with the sum of the corresponding elements in x and y:

Tensor.Add(x, y, result);

In addition to working with individual elements, this library supports various aggregation operations. For example:

var sum = TensorOperations.Sum(data);
var min = TensorOperations.Min(data);
var (min, max) = TensorOperations.MinMax(data);

It also helps find the index of specific elements:

var index = TensorOperations.IndexOfMin(data);

Additionally, it can quickly locate the first element meeting certain criteria:

var value = FirstGreaterThan(data, 0);

And it simplifies finding the index of such elements:

var index = IndexOfFirstGreaterThan(data, 0);

Custom Operations

While NetFabric.Numerics.Tensors provides various primitive operations, combining them might not be efficient. Custom operators can be implemented, allowing the definition of specific operations for each element of the source, while still benefiting from high-performance reusable iteration code.

Credits

This project relies on the following open-source projects:

• .NET
• BenchmarkDotNet
• xUnit

License

This project is licensed under the MIT license. Refer to the LICENSE file for details.