Release 3.1.0

README.md

@@ -36,12 +36,25 @@ For variables `x`, `y`, and `result`, all of type `Span<float>` and of the same
 Tensor.Add(x, y, result);
 ```
 
-Beyond per element operations, the library extends support to aggregations. For a `values` variable of type `Span<float>`, the following code snippet calculates the sum, the minimum and the minimum magnitude of all its elements:
+In addition to per-element operations, the library also supports various aggregation operations, including:
 
 ```csharp
-var sum = Tensor.Sum(values);
-var min = Tensor.Min(values);
-var minMagnitude = Tensor.MinMagnitude(values);
+var sum = TensorOperations.Sum(values);
+var average = TensorOperations.Average(values);
+var product = TensorOperations.Product(values);
+var min = TensorOperations.Min(values);
+var max = TensorOperations.Max(values);
+var (min, max) = TensorOperations.MinMax(values);
+```
+
+The library offers functionality to retrieve the index of the first element meeting a specific condition:
+
+```csharp
+var indexOfEqual = IndexOfEquals(values, 0);
+var indexOfGreaterThan = IndexOfGreaterThan(values, 0);
+var indexOfGreaterThanOrEqual = IndexOfGreaterThanOrEqual(values, 0);
+var indexOfLessThan = IndexOfLessThan(values, 0);
+var indexOfLessThanOrEqual = IndexOfLessThanOrEqual(values, 0);
 ```
 
 ### Custom Operations

docs/articles/Getting-started.md

@@ -35,7 +35,7 @@ var average = TensorOperations.Average(values);
 var product = TensorOperations.Product(values);
 var min = TensorOperations.Min(values);
 var max = TensorOperations.Max(values);
-(var min, var max) = TensorOperations.MinMax(values);
+var (min, max) = TensorOperations.MinMax(values);
 ```
 
 The library offers functionality to retrieve the index of the first element meeting a specific condition:

src/NetFabric.Numerics.Tensors/NetFabric.Numerics.Tensors.csproj

@@ -8,24 +8,18 @@
       This library, based on .NET generic math, provides methods for performing mathematical operations over spans of value types. 
       These operations can be accelerated using SIMD operations supported by the CPU where available.
     </Description>
-    <Version>3.0.0</Version>
+    <Version>3.1.0</Version>
     <PackageIcon>Icon.png</PackageIcon>
     <PackageLicenseFile>LICENSE</PackageLicenseFile>
     <PackageReadmeFile>README.md</PackageReadmeFile>
     <PackageTags>math, performance, intrinsics, simd, span</PackageTags>
     <PackageProjectUrl>https://netfabric.github.io/NetFabric.Numerics.Tensors/</PackageProjectUrl>
     <PackageReleaseNotes>
-- Changed the namespace to NetFabric.Numerics.Tensors.
-- Refactoring of the IAggregationOperation interface.
-- Added support for transform operation on Aggregate.
-- Added AggreagatePropagateNaN.
-- Renamed IGenericBinaryOperation to IBinaryScalarOperation and ApplyGeneric to ApplyScalar.
-- Fixed Min and Max behavior.
-- Added SinCos operations returning the result in two separate spans.
-- Added operations: MinMagnitude, MaxMagnitude, and Sigmoid.
-- Added aggregation operations: Sum, Product, Min, Max, and MinMax.
-- Vectorized Aggregate2D, Aggregate3D, and Aggregate4D.
-- Vectorized operators: ScaleB.
+- Improved performance of Aggregate2D, Aggregate3D, and Aggregate4D.
+- Made the operators public.
+- Added interfaces IUnaryToScalarOperator, IBinaryToScalarOperator, and ITernaryToScalarOperator.
+- Added IndexOfPredicate method.
+- Added operations: IndexOfEquals, IndexOfGreaterThan, IndexOfGreaterThanOrEqual, IndexOfLessThan, and IndexOfLessThanOrEqual.
     </PackageReleaseNotes>
     <RootNamespace>NetFabric.Numerics.Tensors</RootNamespace>
     <PublishRepositoryUrl>true</PublishRepositoryUrl>

src/NetFabric.Numerics.Tensors/README.md

@@ -28,15 +28,25 @@ Given variables `x`, `y`, and `result`, all of type `Span<float>` and of the sam
 TensorOperations.Add(x, y, result);
 ```
 
-Beyond per element operations, the library extends support to aggregations. For a `values` variable of type `Span<float>`, the following code snippet calculates the sum, the minimum and the minimum magnitude of all its elements:
+In addition to per-element operations, the library also supports various aggregation operations, including:
 
 ```csharp
 var sum = TensorOperations.Sum(values);
 var average = TensorOperations.Average(values);
 var product = TensorOperations.Product(values);
 var min = TensorOperations.Min(values);
 var max = TensorOperations.Max(values);
-(var min, var max) = TensorOperations.MinMax(values);
+var (min, max) = TensorOperations.MinMax(values);
+```
+
+The library offers functionality to retrieve the index of the first element meeting a specific condition:
+
+```csharp
+var indexOfEqual = IndexOfEquals(values, 0);
+var indexOfGreaterThan = IndexOfGreaterThan(values, 0);
+var indexOfGreaterThanOrEqual = IndexOfGreaterThanOrEqual(values, 0);
+var indexOfLessThan = IndexOfLessThan(values, 0);
+var indexOfLessThanOrEqual = IndexOfLessThanOrEqual(values, 0);
 ```
 
 Please consult the [documentation](https://netfabric.github.io/NetFabric.Numerics.Tensors/api/NetFabric.Numerics.TensorOperations.html) to get the complete list of pre-defined operations.