Reimplementation of QSARModeling in Rust for performance reasons.

Simply compare running cargo run --release -- 10 -y data/atividades.csv -s ";" (~4 s) vs running python src/main.py (~2 min). The data matrix isn't available on the repo because of Github size limit, feel free to contact me if you want some test files (or just use any training/supervision data).

After cloning the repo, run cargo build --release to build.

Then you can run the program with cargo run -- preceding the options. To see CLI options, run

$ cargo run --release -- --help

Create a new environment and install all dependencies:

$ python -m venv .env
$ source .env/scripts/activate
(.env) $ pip install .
(.env) $ maturin develop

The last command, maturin develop, will build the Rust package and install the Python bindings to the .env environment.

See also: More information about Maturin and building.

Note: Maybe you need to have your Python interpreter set on PYO3_PYTHON environment variable, e.g. export PYO3_PYTHON = ~/qsarmodelingrs/.env/bin/python3.

Heliton Martins [email protected]