Friends of Friends Algorithm

Disclaimer: WORK IN PROGRESS!!

This project uses the Friends-of-friends (FoF) algorithm to search for high redshift (0.5 < z < 2) galaxy clusters within the COSMOS survey.

This code base was developed for and used in my undergraduate thesis:

Detection and Analysis of High Redshift Galaxy Clusters

Completed: May 2021

About

• Separations and nearest neighbour lookups are sped up with GrisPy bubble neighbors search.
• Uses 3-sigma method to remove interlopers.
• Calculates mass, velocity dispersion and projected radius with the virial mass estimator.
• Performs catalog matching with other notable COSMOS galaxy cluster catalogs.
  • F. Bellagamba et al., 2011
  • George et al., 2011
  • Z. L. Wen & J. L. Han, 2011
  • Ilona K. Söchting et al., 2012
• Parameters chosen are empirically determined

Dependencies

• Python 3.6 or later
• Popular STEM libraries including: Numpy, Scipy
• Data Processing libraries: Pandas, Matplotlib
• Astropy
• GrisPy

Usage

1. Pull this repository with

git pull https://github.com/kennethcheo/FoF.git

2. Obtain your dataset in your desired format (FITS, csv etc.)

   • This project mainly uses Pandas' dataframes to hold data.
   • Refer to data_processing.py for conversion of FITS to DF

3. Ensure galaxy photometric or spectroscopic dataset with the relevant properties: RA, DEC, z, Absolute Magnitude, Galaxy ID

4. Run pipeline with pipeline.py by commenting out relevant parts (or using it as a guide)

5. Remember to change the default parameters in params.py

6. Run dataset through the pipeline and save them to the appropriate location

   • Large datasets (> 1 million points) might take awhile.

Parameters

• max_velocity: Maximum relative velocity of member galaxies to cluster center.
• linking_length_factor: Constant of proportionality in determining linking length, default is 0.2 for dark matter halos.
• virial_radius: Maximum separation of member galaxies from to cluster center.
• richness, R: Minimum number of galaxies to form a cluster.
• overdensity, D: Cluster density comparison with background density.