pyROGER package

Documentation Status

Welcome to the pyROGER package, where you will be able to make an automatic dynamical classification of galaxies around clusters.

Installation

You can install pyROGER by doing:

pip install pyROGER

Or install the last version downloading the source:

git clone https://github.com/Martindelosrios/pyROGER.git

And then, in the pyROGER directory just doing:

pip install .

Requirements

For running pyroger you will need:

  • numpy

  • scikit-learn (If you want to use the pre-loaded models, you will need to install scikit-learn==1.3.0)

  • mlxtend

  • pandas

  • joblib

Quick Start

First you can import the roger library and the roger models…

from pyROGER import roger
from pyROGER import models

Then you can check the available models…

models.list_saved_models()

For uploading a model just give the path to the saved .joblib file

models.HighMassRoger1.train(path_to_saved_model= [PATH_TO_MODEL/HighMassRoger1_KNN.joblib',
                                                  PATH_TO_MODEL/HighMassRoger1_RF.joblib',
                                                  PATH_TO_MODEL/HighMassRoger1_SVM.joblib'])

To check that the model is ready to be used you can do:

models.HighMassRoger1

and you should see ‘Ready to use’ after some comments about the model.

Finally, for using the model you just do:

# For predicting the classes:
pred_class = models.HighMassRoger1.predict_class(data[:,(0,1)], 0)

# For predicting probabilities:
pred_prob = models.HighMassRoger1.predict_prob(data[:,(0,1)], 0)

Tutorials

In Examples you can find some examples to start playing with pyROGER

Citation

If you use pyROGER in your paper please cite https://arxiv.org/abs/2010.11959

@ARTICLE{2021MNRAS.500.1784D,
       author = {{de los Rios}, Mart{\'\i}n and {Mart{\'\i}nez}, H{\'e}ctor J. and {Coenda}, Valeria and {Muriel}, Hern{\'a}n and {Ruiz}, Andr{\'e}s N. and {Vega-Mart{\'\i}nez}, Cristian A. and {Cora}, Sof{\'\i}a A.},
        title = "{ROGER: Reconstructing orbits of galaxies in extreme regions using machine learning techniques}",
      journal = {\mnras},
     keywords = {methods: analytical, methods: numerical, galaxies: clusters: general, galaxies: kinematics and dynamics, Astrophysics - Astrophysics of Galaxies},
         year = 2021,
        month = jan,
       volume = {500},
       number = {2},
        pages = {1784-1794},
          doi = {10.1093/mnras/staa3339},
archivePrefix = {arXiv},
       eprint = {2010.11959},
 primaryClass = {astro-ph.GA},
       adsurl = {https://ui.adsabs.harvard.edu/abs/2021MNRAS.500.1784D},
      adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}
A lot of people is already using pyROGER, you can check their papers here!(https://scixplorer.org/abs/2021MNRAS.500.1784D/citations?p=1)

pyROGER documentation

Contents: