The notebooks in this book are organised by concept, not by teaching week. Start with the relevant core chapter, then use a notebook to explore the methods and interpretation. Some applied notebooks need separate datasets; check each notebook’s data requirements first.
1. Your Python environment¶
Use a recent Python 3 environment in JupyterLab, Jupyter Notebook or Google Colab. Google Colab is optional: the public notebooks should not require access to the GE5230 course Google Drive.
Install only the packages needed by your chosen notebook. A common starting set is numpy, pandas, matplotlib, scipy, geopandas, shapely, scikit-learn, mapclassify, libpysal, and esda. Some notebooks also use pingouin or additional visualisation packages.
2. Check your environment¶
import sys
import importlib.util
print('Python:', sys.version.split()[0])
for package in ('numpy', 'pandas', 'matplotlib', 'scipy', 'geopandas', 'sklearn', 'libpysal', 'esda'):
available = importlib.util.find_spec(package) is not None
print(f'{package}: {"installed" if available else "not installed"}')
3. Finding the data¶
Look for a Data and requirements section at the start of each notebook. Paths should be relative to the repository or explicitly configurable. The Part 3 catalogue connects notebooks to related chapters and the data catalogue.
If a notebook requires an external dataset, obtain it from its stated source and follow its setup instructions. Do not assume that private course Drive folders are available.
4. Reading results¶
Record the observation unit, study window or areal units, software versions where material, assumptions, and uncertainty. If you change a parameter, ask whether the analytical question has changed as well.