Melbourne datathon notebooks and data processing scripts.
Manual Setup Steps:
- Visit https://conda.io/miniconda.html and download the Python 3.6 installer
- Run the installer, but don't set up the environment. Instead follow these instructions.
Automated Setup Scripts:
-
'sh basic.sh' <-- this will create the "dthon" Python environment with the basic libraries to load and visualise the data
-
'sh advanced.sh' <-- this will add more comple scientific and data science libraries for statistical modelling
Using the Environment:
- Run 'source activate dthon'. This will start you using the environment we just set up.
- Run 'jupyter notebook'. This will launch a web-based Python environment.
Working with the example notebooks:
Tennessee, Ioanna and Nathan will maintain a set of basic example notebooks in the directory 'notebooks/examples/' to demonstrate basic techniques.
About the directory layout:
-
scripts/ <-- command-line tools to execute a process
-
src/ <-- handy methods and functions for re-use
-
notebooks/
-
examples/ <-- showing based techniques
-
<name>/ <-- personal working directories
-
submissions/
-
<name>/ <-- put your submission files from your experiments here
Suggestions for good work practises:
- Whenever you make a submission that's an improvement, create a copy of that notebook and put a version number on it.