Elastic Net–based Indicator Selection for Nowcasting German GDP

Evidence from Unrestricted MIDAS Models

Master Thesis by Johannes Schäfer

---

The project implements a two-step nowcasting framework using Elastic Net–based indicator selection and Unrestricted MIDAS models.

---

Getting Started

To set up the required Python environment, run:

conda env create -f conda_env.yml
conda activate DataAnalysis
Then open the file:
/Code/main.py
for detailed run instructions on how to execute the scripts and reproduce the results.

Author
Johannes Schäfer
5577001