joclement/master_thesis

Implementation of my master's thesis

Implementation for Master Thesis

Windows

Check e.g. the official python docs to install Python and make a command line ready.

Installation & Start

On a command line install with the following command when the wheel is downloaded:

python -m pip install FILEPATH_TO_WHEEL

Then start the GUI by:

python -m jc-thesis-code

For future development

This is the code to the master's thesis from Joris Clement with the title "Evaluation of Classification Algorithms for Partial Discharge Diagnosis in Gas-Insulated HVDC Systems". The code is written in Python. Poetry is used for the management of the dependencies and to execute the scripts. Check the Poetry documentation on how to install it. When Poetry is installed execute the following:

poetry install

With nox the tests and other checks can be run. Therefore nox needs to be installed. See the documentation here. Then just type nox on the command line to run the tests and checks.

After the dependencies are installed and the scripts can be executed. Run for e.g.

poetry run classify config/fingerprints.yml

to train and validate some of the feature-based models. The data is needed to do that. The data is contained in the submodule data and not in this repo directly due to its disk usage of more than 5GB. It was further required to keep the data private. Example data can be found ./testdata/.