Sukasa is a Python-based web application that analyses current house prices to detect and report price imperfections. This application provides prospective house-buyers a way to investigate whether the houses they are purchasing are incorrectly valued based on social, economic, and geographic characteristics such as current wage levels, the price of similar houses in neighbouring areas, nearby amenities such as schools and shops, and the estate agent selling the house. The aim is to give more information — and therefore sovereignty — to potential consumers, or even just to help current homeowners overcome buyers remorse.
The Python web application uses Django as the backend framework and Angular as the frontend framework. Documentation of the RESTful API service is handled by Swagger. Development is done within a Docker environment and run on Google Kubernetes. We use ElasticSearch as our main database and Kibana as the database visualisation GUI. We use Redis as the in-memory data store to reduce strain on the main database. R is the language of choice for any statistical research.
We use Docker in the development of Sukasa to make it easy to build, run and share. We have two Dockerfile's: one the builds the Python backend and one that builds the Angular GUI. Both components of the application can be built from the build.sh file:
cd Sukasa
bash build.sh all
The build.sh file has multiple options to choose from.
The build.sh file also coordinates the running of the application. Specifically, the application can be run with the following command:
bash build.sh up
Likewise, the application can be brought down with bash build.sh down.
By default, the main Sukasa application is accessible on port 3000, the API and backend components are on port 5000, the ElasticSearch database is run on port 9200, Kibana is on port 5601 and Redis is run on port 6379. The docker-compose.yml can be altered to allow these services to run on different ports. Swagger API documentation can be found at localhost:5000/swagger/.
Environmental variables for the application can be changed in the docker-compose.yml file.
The best way to troubleshoot or ask for a new feature or enhancement is to create a Github issue. However, if you have any further questions you can contact me directly.