The rust-ml group is still in its infancy, but these are already available:

• linear decision trees
• non-linear manifolds
• linear regression
• a Torch model, for example CIFA-10 dataset with CNN

We should also have an introduction to, not only the algorithms, but also the tools we are using:

• How to use the ndarray crate and understand type signatures of ArrayBase
• Give examples for ndarray-stats
• Integration of gnuplot into the workflow

This list is far from complete and will grow with time. If you know any algorithm, you want to add, please feel free to contribute.

/// Convert input string to 24 character hash
pub fn hash(input: &str) -> String {
let mut sh = Sha256::new();
sh.input(input.as_bytes());
let mut out = format!("{:x}", sh.result());
out.truncate(24);
out
}

Does not build, input and result are not registered apis

changed to
/// Convert input string to 24 character hash
pub fn hash(input: &str) -> String {
let mut sh = Sha256::new();
sh.update(input.as_bytes());
let mut out = format!("{:x}", sh.finalize());
out.truncate(24);
out
}

Now that linfa 0.6 has been released, which does not require a BLAS dependency, it should be much easier for users to get started with using it. This book should be updated to the 0.6 release, and specify the 0.3.1 release of plottersfor generating plots.

We should think of a way to add multiple languages in the book. The easiest option would add a separate mdbook for each language and publish each of them separately.

Relevant issue rust-lang/mdBook#5

Per the Rust Embedded Book, it looks like it's common to include a license for the prose sections of a book as well. I believe that they use the CC-BY-SA 4.0 license, terms of which can be seen here.

We should build and publish the book automatically to rust-ml.github.io/book on each push:

• add small CI system which builds the book
• add gh branch with artifacts

At the moment this code snippets

user@computer:~/rust-ml/book$ cd code/
user@computer:~/rust-ml/book/code$ cargo run --release --example kmeans

is parsed with the dollar symbols as inline equations.