-
While a bit niche in the tensor space, index views are an undeniably useful feature.
[xtensor has support for index views](https://xtensor.readthedocs.io/en/latest/view.html#index-views), though re…
-
Basic idea is now in place, but implementation can still be tricky...
Will implement for AnalogSignalArray first...
-
Inspired by mscha@IRC
## The Problem
An array can be indexed many different ways, for example, using the Whatever star, or Ranges that end with a Whatever star. However, when such a Range is use…
-
``` nemerle
def a = array(2, 2);
WriteLine(a[(0, 1)]);
```
> error: expected (int- \* int-), got int in array index: the types int and (int- \* int-) are not compatible [simple unify]
The error does…
-
## 🚀 Feature
Let [`index_select`](https://pytorch.org/docs/stable/torch.html#torch.index_select) work with a multidimensional `index`.
## Motivation
Index the input tensor along a given dimensi…
-
## 🚀 Feature
Allow multidimensional index in `index_add_`
## Motivation
`index_add_` is similar to `scatter_add_` with the index applied on RHS rather than LHS. Unfortunately, currently `scatter…
-
If `A` is a Matrix, I can (of course) access its elements by either linear or cartesian indices.
However, constructing a reference to an entry of `A` is possible only by linear indexing:
```julia
…
-
Currently `clad::array` and `clad::array_ref` only supports single dimensional arrays. If this feature is added clad will be able to differentiate multidimensional arrays. Possible design decisions in…
-
Ceylon doesn't have true multidimensional lists, but it is possible to work around that using a list of lists as in Java. This works out just perfectly for multidimensional tuples:
```ceylon
value…
-
## Description
NDArray#get fails with pick index for multidimensional arrays:
Executing code:
```
NDArray target = manager.arange(6).reshape(3, 2);
NDArray index = manager.create(new long[] {0, 2…