bokeh_server_example

Minimal example of a bokeh dashboard, mainly to find out how this can be hosted

Installing dependencies

• Install Miniconda or Anaconda
• Open a terminal in the root folder of this repo and type
  • conda env create -f environment.yaml
  • conda activate covid-dashboard

Running the dashboard

Type python run_dashboard.py in a terminal in the root folder of this repo.

This will open a browser window with the minimal dashboard. If you want to specify the port, do this at the end of run_dashboard.py.

Background information

The dashboard runs on a bokeh server. The detailed server architecture is explained here.

All the data needed in the dashboards is loaded once before the Bokeh server is started. It is stored in simple .csv and .json files. The session specific data is managed by the bokeh server. There is no need for a database to handle user requests.

The information shown in the plots is pre-calculated. Thus the main operation we perform are dictionary lookups, which are very fast. The whole application is probably not very resource intensive.

Differences between the minimal example later version

• There will be a proper command line interface with options to specify the port
• The data files will be larger, but probably not large. Probably less than 1 MB.