mstubna/hands-on-ml-mnist

This repo contains scripts that work through the examples in Chapter 3 of the Hands On Machine Learning book.

Scripts

• visualize_data.py -- generates data visualizations

Initial project setup

1. If necessary, install conda or miniconda
2. Clone the repo
3. From a terminal shell, run conda env create --file environment.yml, then conda activate hands-on-ml-housing
4. Create a data folder and download the housing data file and extract it to the data folder.
5. You should now be able to run the scripts, for example by, python visualize_data.py

Updating dependencies

1. To update your local dependencies to match the versions in the environment.yml file, run conda env update --file environment.yml

Installing new dependencies

1. Install new package(s) by running conda install <package_1> <package_2>.
2. To update outdated packages, run conda update --all, which will show a list of packages that will be updated. Reply N to just see the list or Y to install the new packages.
3. After updating packages, run conda env export > environment.yml to update the environment.yml file; commit the changes.