Skip to content

Latest commit

 

History

History
146 lines (97 loc) · 3.52 KB

README.md

File metadata and controls

146 lines (97 loc) · 3.52 KB

AI Workshop

Notebooks 🪐

A collection of Jupyter notebooks for various ML tasks.

Notebooks ToC

Don't Know Where to Start?

Why Not Crab?

Notebooks Viewer

If you're having trouble viewing the notebooks in GitHub, try using nbviewer.

https://nbviewer.org/github/ahester57/ai_workshop/blob/master/notebooks/time_for_crab/0-eda/eda.ipynb


Development

Create a Virtual Environment

Create the virtual environment with the following commands:

$ mkdir -p ~/dev
$ python -m venv ~/dev/.venv
$ source ~/dev/.venv/Scripts/activate
$ echo "Confirm you're using the correct python with: "
$ which python
=> Should show path to your .venv

Alternatively, you can use the provided bash scripts to enter into it with source-ry.

$ ./create-venv.sh
$ source source-me-to-activate-venv.sh

Install Development Dependencies

(.venv)
$ pip install -e ".[dev]"

Install with Optional Dependencies

In the above, ".[dev]" means you want to install the optional packages under 'dev' in the project pyproject.toml.
There is also a visual mode driven by networkx and matplotlib. To install dev tools plus visual mode, use this pip install command instead:

(.venv)
$ pip install -e ".[dev,jupyter]"

See this reference for more info about installing Python packages.

Running

Now you can run the application from within the virtual environment:

(.venv)
$ aiw

Test Suite

Run tests:

(.venv)
$ python -m pytest -v tests/

Documentation

Contents

See the docs directory for more.


Execution

Once installed, the application can be run via the command line. Inside your virtual environment:

(.venv)
$ aiw --help
usage: aiw [-h] [-c CONFIG] [-v] [-w WARN] {anneal,svm,convolve} ...

options:
  -h, --help            show this help message and exit
  -c CONFIG, --config CONFIG
                        config file [etc/config.toml]
  -v, --version         print version and exit
  -w WARN, --warn WARN  logger warning level [WARN]

subcommands:
  {anneal,svm,convolve}
(.venv-ai)

Configuration

The application uses TOML files for configuration. Configuration supports runtime parameter substitution via a shell-like variable syntax, i.e. var = ${VALUE}. CLI invocation will use the current environment for parameter substitution, which makes it simple to pass host-specific values to the application without needing to change the config file for every installation. Config file is located in /etc.

    logging = "INFO"

Logging

The application uses standard Python logging. All logging is to STDERR, and the logging level can be set via the config file or on the command line. Find the code in the core module.