Morphological Classification of Magnetar Bursts

Codebase outline

Click on their name to see the source codes.

Event handling, visualization, data extraction

• iddb.csv : The table where event names and their corresponding ids are stored.
• Event/Event.py : The object-oriented source code for analyzing event records, extracting features, visualizing nearly everything about any event on their own. It's a custom package!
• Event/test.py : Code required for testing the functionality of Event.py
• features.py : Handles event files, acts as a bridge between Event.py and event record files, writes every event's features to a file named export_data.csv. A total of 221 pulse features is extracted from 101 event records.

Data analysis, model training, further visualization of clusters/events

• clusters.csv : The data file including KMeans and DBSCAN clusterings of all the events.
• model/kmeans.py : A code that applies various analysis techniques and machine learning algorithms to the extracted event data.
  • Analysis Techniques: Principal Component Variance Analysis, Feature importance sorting, Silhouette Scores of clusterings
  • Machine Learning Algorithms: K-means clustering, DBSCAN clustering, Hierarchial Clustering
• model/kmeans1.py : A statistical analysis and 3D visualization code for understanding the behaviours of features and visualizing 3D view of clusters.
• model/simulation.py : A simulation for understanding how varying specific features change explained variance ratio, also checks for correlations between Principal Components and features, and visualizes a BiPlot.
• model/test3ax.py : A test for visualizing 5D/4D/3D clusters in 3D space.

Running the code

requirements.txt file includes all the necessary packages to develop and run this project properly. The main packages used in this project are:

matplotlib
pandas
sklearn
scipy
numpy
seaborn
kneed

First clone the repository:

git clone https://github.com/Abtaha/PROJ201.git

Then move into the repository and install the required packages:

pip install -r requirements.txt

Then run the following python files from the root directory of the repository to see the visualizations and analysis:

model/kmeans.py
model/simulation.py