Demographic Modeling is opinionated tooling for creating geographic factors for machine learning.
No, this set of tooling written in Python is not going to have a political debate with you. Rather, while flexible enough to be used in a variety of ways, this tooling provides a clear way to perform analysis. This enables you to get started and be productive as quickly as possible.
There is a reason why DataFrames are so ubiqutous across all data science data tools. DataFrames, tabular data with built in functionality, they make working with data expendetially easier. In this project, almost everything takes input data as a Pandas DataFrame, and if the data is geographic, spatial, it is an Esri Spatially Enabled DataFrame.
Frequently one of the most difficult things about working with spatial data is knowing where it is. Spatial reference, many times referred to as the the projection, is how we know where things are located on the face of the planet. The most commonly used spatial reference is WGS84, the longitude and latitude used by GPS systems, including those in smart phones.
However, for those of us who spend our lives working with geographic data, how we know where something is, or is not, is not as simple as it seems. Data can come in a variety of spatial references, and to make working with data coming together potentially in differing spatial references easier, this package converts (projects) data into WGS84 by default so the data is easier to deal with and the coordinates are easily recognized by most people, geographers and data scientists alike.
From the project directory, create an environment with all dependencies installed and linked.
> make env
If you want to manually activate the environment, this can be accomplished with the commmand.
> conda activate demographic-modeling
An example workflow can be found in the Jupyter notebooks in the ./notebooks directory of the project. If you want
to get started quickly, you can use the following command. It activates the environment and starts Jupyter Lab in
one consolidated step for you.
> make jupyter
Project based on the cookiecutter GeoAI project template. This template, in turn, is simply an extension and light modification of the cookiecutter data science project template. #cookiecutterdatascience