dylantartarini1996 / energy_prices_monitoring_ita Goto Github PK

We want to make the web app as accessible as possible for everyone

• color themes (for people who are colorblind)
• characters fonts (for people with dislexya)
• possibilities for visually impaired people

Create new Forecaster classes out of different forecasting algos:

• create the new classes
• make them available to users in app pages (for instance, give the ability to choose the algo to use..?)
• start with ARIMA models

Use the homepage of the Streamlit App to give out information about the project and what are the available features in the pages

Links to data for

• Fuels
• Electricity
  changed. Need to adjust the code accordingly

Create UnitTests for

• forecaster class

Create IntegrationTests for

• data loading
• mlflow

In the Fuels page, we are dealing with 3 time-series. Adapt the train/predict codebase from the other pages to be able to

• choose the time series of interest (target)
• visualize all ts in the plot until the choice is made, then plot only the target
• train the model
• predict/forecast

Some functionalities in the pages could be defined only once in a utils.py module and loaded from there