Classification of Fermi LAT Gamma-ray Sources from the FL8Y Catalog using Machine Learning Techniques

To use the code, follow the instructions below.

Prerequisites

1. Clone all the file to your local computer.

2. Make sure you have the required libraries installed for your Python 3 (NumPy, Pandas, scikit-learn and pickle [pickle is installed for Python 3 by default]).

AGN-PSR Classification

1. For AGN-PSR Classification, run

python3 AGN_PSR_Training.py

The overall training results will be stored in AGN_PSR_result.txt.

2. To classify the whole catalog, run

python3 AGN_PSR_Result.py

The classified catalog will be stored in results.csv. The list of PSR for the two different models (90.0%-up-model and top-3-model) will be stored in psr_candidates_90_up_model.txt and psr_candidates_top_3_model.txt.

YNG-MSP Classification

1. (Optional) If you don't have the file psr_list.txt for the list of YNG/MSP, run

g++ gen_psr_list.cpp -o gen_psr_list
./gen_psr_list

2. For YNG-MSP Classification, run

python3 YNG_MSP_Training.py

The overall training results will be stored in YNG_MSP_result.txt.

3. To classify the PSR candidates (Make sure you have run the AGN-PSR Classification for the list of PSR candidates), run

python3 YNG_MSP_Result.py

The classified catalog will be stored in results_msp.csv. The list of MSP for the two different models (90.0%-up-model and top-3-model) will be stored in msp_candidates_90_up_model.txt and msp_candidates_top_3_model.txt.