You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have been hand-writing a pile of generic kernels for numerical arrays and wanted to know what would be required to contribute them upstream.
So far I have implementations for:
public static Array Filter(Array array, BooleanArray mask)
public static BooleanArray Equal<T>(PrimitiveArray<T> lhs, T rhs) where T: struct, INumber<T>
public static BooleanArray Equal<T>(PrimitiveArray<T> lhs, PrimitiveArray<T> rhs) where T: struct, INumber<T>
public static Array Take(Array array, IList<int> indices)
public static Int64Array CastInt64<T>(PrimitiveArray<T> array) where T: struct, INumber<T>
public static Int32Array CastInt32<T>(PrimitiveArray<T> array) where T : struct, INumber<T>
public static FloatArray CastFloat<T>(PrimitiveArray<T> array) where T : struct, INumber<T>
public static DoubleArray CastDouble<T>(PrimitiveArray<T> array) where T : struct, INumber<T>
I also have a generic visitor for converting Large* arrays to their regular counterparts so that they may work with the other machinery in the library. I am less confident that is useful to others, or that it is truly correct.
Is there a particular pattern you would want these to follow or is the notion not suitable for the .NET ecosystem? I am less familiar with .NET as an ecosystem.
Thanks for your interest! This sounds very useful. I would suggest creating a new project Apache.Arrow.Operations for these kernels and perhaps we can iterate on it a bit before publishing it and locking in the APIs.
Activity
CurtHagenlocher commentedon Feb 3, 2026
Thanks for your interest! This sounds very useful. I would suggest creating a new project Apache.Arrow.Operations for these kernels and perhaps we can iterate on it a bit before publishing it and locking in the APIs.