Accept owned arrays in `concatenate` and `stack`

未关闭
#1,591 2 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

还没有人认领这个 Issue。

评估

难度
3/5
预计耗时
1-2 天
新手友好度
45/100
Issue 类型
功能
描述清晰度
基本清楚
活跃度
冷清
技术栈
rust
领域
data

调研方向

Start by reading the current concatenate and stack entry points and their signatures. Check how they accept views, how owned arrays would fit, and what happens for an empty slice. Done means the proposed API works for owned arrays and views without leaving the documented inference concern unresolved.

由索引模型根据 Issue 内容生成。

描述

breaking-change enhancement

It would be convenient to be able to call concatenate or stack with a slice of owned arrays or array views, rather than just views currently.

So for concatenate, this:

let arrays = vec![a, b, c]; // more realistically, this would be returned by a function
let arrays = arrays.iter().map(|a| a.view()).collect::<Vec<_>>();
let res = ndarray::concatenate(Axis(2), &arrays);

would become:

let arrays = vec![a, b, c];
let res = ndarray::concatenate(Axis(2), &arrays);

This could be done by changing the signature of concatenate to the following, which would allow either views or owned arrays:

-pub fn concatenate<A, D>(axis: Axis, arrays: &[ArrayView<A, D>]) -> Result<Array<A, D>, ShapeError>
+pub fn concatenate<S, D, A>(axis: Axis, arrays: &[ArrayBase<S, D, A>]) -> Result<Array<A, D>, ShapeError>
 where
+    S: Data<Elem = A>,
     A: Clone,
     D: RemoveAxis,
 {

This is technically a breaking API change though, at least in the niche case of passing an empty slice as a parameter (which causes a type inference error). Before I open an unsolicited PR, would there be interest in this change?

主要语言
Rust
星标
4.3k
派生
391
30 天内没有已合并 PR

贡献指南

这个仓库没有索引到贡献指南

从这里开始

  1. 先读完整个 Issue，再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库，在一个分支上完成修改。
  4. 提交 Pull Request，并在描述里引用这个 Issue 编号。

rust-ndarray/ndarray 的其他 Issue

查看 rust-ndarray/ndarray 的全部 Issue

相似的 Issue

更多 Rust Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。