Pavel Gorgulov

As I see it, the most critical points here are: 1. Verify the iteration over ndarray. Ensure that a zero value does not break array traversal since it depends on...

There is a `pow` function, but it raises the entire matrix to the power. It makes sense to add the function you described to the library

`mk.linalg.pow(ndarray, n)` - https://kotlin.github.io/multik/multik-core/org.jetbrains.kotlinx.multik.api.linalg/-lin-alg/pow.html?query=abstract%20fun%20%3CT%20:%20Number%3E%20pow(mat:%20MultiArray%3CT,%20D2%3E,%20n:%20Int):%20NDArray%3CT,%20D2%3E

> Any plans to support power to float/double besides just int? Yes, it is planned > What about in-place? I'm avoiding using inplace operations for now, as it creates a...

Thank you very much for the additional information. I have reproduced the error. It indeed occurs with the native implementation, which is selected by default for tensor multiplication. As far...

Unfortunately, there’s no workaround * for Android, you can use the JVM implementation of reading .npy files. * for iOS, you’ll need to implement it manually. The JVM library used...

Hi Yes, there is There's `sl.bounds`, and also `Slice` class Negative indices are also supported. In my opinion, this is not the best API for boundaries, so I'm considering improving...