rust-numpy VS docs.rs

On the contrary, it can use and interface with numpy quite easily: https://github.com/PyO3/rust-numpy

Once I figured out the right tools, it was easy. Its just "maturin new". It automatically converts python floats and strings. Numpy arrays come through as a special Pyarray type, that you need to unwrap, but that's just one builtin function. Using pyo3, maturin and numpy, https://github.com/PyO3/rust-numpy it's fairly easy.

If I'm understanding this documentation correctly then you may be able to pass the numpy array directly with func(df['col'].to_numpy) which may save some conversion.

Otherwise, though, Rust is an excellent choice. The many advantages of Rust (great package manager, memory safety, modern language features, ...) are already well documented so I won't repeat them here. Specifically for writing Python libraries, check out PyO3, maturin, and rust-numpy, which allow for seamless integration with the Python scientific computing ecosystem. Dockerizing/packaging is a non-issue, with the aforementioned libraries you can easily publish Rust libraries as pip packages or compile them from source as part of your docker build. We have several successful production deployments of Rust code at OpenAI, and I have personally found it to be a joy to work with.

Integration with numpy uses the rust-numpy crate: Example of method that accepts numpy arrays as arguments Example of a method that returns a numpy array to Python (this performs a copy, there ought to be a way to avoid it but the current implementation has been plenty fast for my use case so far)

This example maybe helpful.

Given that it's via pyO3, you could even pass the numpy arrays using https://github.com/PyO3/rust-numpy and get ndarrays at the other side.

Same no copy, slightly more user friendly approach.

Further criticism of the actual approach - even if we didn't do zero copy, there's no preallocation for the vector despite the size being known upfront, and nested vectors are very slow by default.

So you could speed up the entire thing by passing it to ndarray, and then running a single call to sum over the 2D array you'd find at the other end. (https://docs.rs/ndarray/0.15.1/ndarray/struct.ArrayBase.html...)

I believe converting between pandas Series (e.g. columns) and numpy ndarrays can be pretty cheap, right? Once they're in that format, you can use rust to work directly on the numpy memory buffer with rust-numpy. Otherwise, feather is a format designed for IPC of columnar data; pyarrow is in pandas (might be an optional dependency) and may be pretty quick for that, and rust has an arrow implementation too.

https://github.com/PyO3/rust-numpy

Working in combination with CodeWhisperer in your IDE, you can send whole code sections to Amazon Q and ask for an explanation of what the selected code does. To show how this works, we open up the file.rs file cloned from this GitHub repository. This is part of an open source project to host documentation of crates for the Rust Programming Language, which is a language we are not familiar with.

Looks like a great initiative – I wish there was a reliable TS/JS equivalent of https://docs.rs (even considering rustdoc's deficiencies[1]).

I went through this exercise recently and so far my experience with trying to produce documentation from a somewhat convoluted TS codebase[2] has been disappointing. I would claim it's a consequence of the library's public (user-facing) API substantially differing from how the actual implementation is structured.

Typedoc produces bad results for that codebase so sphinx-js, which I wanted to use, doesn't have much to work with. I ultimately documented things by hand, for now, the way the API is meant to be used by the user.

Compare:

https://ts-results-es.readthedocs.io/en/latest/reference/api...

vs

https://tsdocs.dev/docs/ts-results-es/4.1.0-alpha.1/index.ht...

Rust Search Extension adds a section on docs.rs menubar which lists the features of a crate in a nice and easy to access format.

📝 Note: At the time of writing this post, I couldn't really locate the init function docs.rs documentation. It didn't seem easily accessible through any of the current HAL implementation documentation. Nevertheless, I reached the signature of the function through the source here.

Rust Crates are a Game-Changer 🎮:The ease of releasing a crate with `cargo publish` and the convenience of rolling out new versions amazed me. The auto-generated docs on Docs.rs. is an amazing tool, especially with docstring formatting. Doc tests serve as a two-fold tool for documenting the code and ensuring it's up-to-date.

I've found I manually type out certain subsets of URLs where possible[0], maybe that's subconsciously associated with my impression that Google Search results have gotten worse and worse over the years.

[0] https://en.wiktionary.org/wiki/ and https://docs.rs/ come to mind.

docs.rs also downloads you crate automatically to generate docs and I would guess lib.rs does something similar

There are also a few key sites you'll want to keep in your back pocket at all times: - The Standard Library Documentation has complete documentation for every std library function in Rust - crates.io is a repository for all third-party packages, and docs.rs has human-readable documentation for the overwhelming majority of them - The Rust Cookbook has some code examples for common tasks you may need to perform - Make sure you are using clippy, which is available through Rustup and can be run with cargo clippy as a replacement to cargo check, it adds additional lints for your Rust code and is very helpful for teaching many of the best practices

OTOH, source-code-generated-docs normalize how code docs are, like the rust docs.rs paradigm, so it sort of forces or encourages package creators/maintainers to write docs.