Members :
Jake Wagoner; u6048387; jakew@sci.utah.edu
Sudhanva Manjunath Athreya; u1529299; u1529299@utah.edu
Tin Vo; u1289616; u1289616@utah.edu
Install Docker execute the following commands.
Frontend is made using Vue and Vuetify for some styling of common components, along with D3. Vite was used for building/running the frontend.
The geojson used for the map was obtained from here: https://gist.github.com/tomshanley/a006f9c6c4b30835137ead7574a51d17#file-utah-districts-geojson
The code in the /web folder uses the Vue3 framework for componentization and responsiveness. There are some components that come from Vuetify, a Vue plugin which utilizes Material UI to pre-style common components. Custom components can be found in /web/components. This includes the VisContainer, all visualizations, sidebars, etc. All frontend code is written by us, leveraging these frameworks/plugins.
The code in the /api folder is our code (using fastapi) which pulls earthquake data from the website given below.
View the screencast here: https://www.youtube.com/watch?v=SErroFogMmk
An API is used to grab data from: https://earthquake.usgs.gov/earthquakes/map/
The range of the data is taken by drawing a rectangle around Utah on the map.
Peak Performance can be achieved on an M1 mac or newer, or a GPU workstation. Read more here : ggml
Provide exec rights to script.sh which is present in the api directory
Executing script.sh should start the LLM. The interactions with the LLM will be handled by the API.
cd api
docker build -t viz_api .
docker run -d --name api_container -p 80:80 viz_api
cd web
docker build -t viz_ui .
docker run -d --name ui_container -p 5173:5173 viz_ui
Open http://localhost:5173/ once both the containers are up.