Energy Prices Monitoring ITA

Project Description

Streamlit Application built for monitoring energy prices in Italy across time.
Italy is not self-sufficient in terms of energy provisioning, but many of its companies are energy-intensive and need to look very carefully at their electricity bill, as well as their transportation/delivery costs; their heating/air conditioning needs and prices associated with them are also not to be underestimated in a period of economic crisis.

This Demo aims to showcase prices for a list of commodities, as well as a Machine Learning model to forecast their evolution; please note that forecast results are NOT a guarantee that market prices will eventually exactly turn out that way, and are only presented as orientative values.

Data Collection

Data for this application are downloaded from a variety of sources; they include

• ⛽️ Fuel prices from Ministero dell'Ambiente e della Sicurezza Energetica;
• ⚡️ Electricity prices from GME
• 🔥 Gas Prices from the TTF index, obtained through the Yahoo Finance API.

Forecasting Algoryhtm(s)

Forecasting for all utilities/fuels is done via Prophet algorithm:

• the default algorithm is the Prophet model developed by Meta;
• model training is monitored and tracked using MLflow;

Local Usage

To run the Streamlit App on your local environment go on your terminal:

1. install the requirements

    pip install -r requirements.txt.

2. run

    streamlit run Home.py

3. experiment tracking is done via mlflow. Once you launch an experiment (i.e. training a model) you can run

    mlflow server

to visualize the mlflow UI and see all the experiment runs for all models, their metrics and so on.