Mulguisin

The Mulguisin(MGS) is a cluster/group finding algorithm in galaxy data. It can find group/cluster structure from the galaxy data and also provide topological informations.

It has been developed by InKyu Park to find jet structure for the LHC experiment. Now, he modify the Mulguisin that can be applied to galaxy data.

Install

The MGS can be install via pip install:

pip install Mulguisin

Source code

The latest version of MGS can be download with:

git clone https://github.com/youngju20/Mulguisin.git

Basic usage

>>> import Mulguisin
>>> import numpy as np
>>> data = np.random.random((10,2))
>>> boundaries = [0,1,0,1]
>>> Rcut = 0.1
>>> mgs_init = Mulguisin.mulguisin_type('voronoi')
>>> MGS = mgs_init(Rcut,data[:,0],data[:,1],boundaries=boundaries)
>>> Nmgs, imgs, clg, clm, cng = MGS.get_mgs()
>>> Nmgs
9

Plot the MGS

• The label contains information about which data belongs to the corresponding MGS. The label is (len(data), ) array and each value in the label indicates the each MGS group.
  e.g.) label = [0, 0, 0, 1, 2]. The first data point is belongs to the first largest MGS group.

>>> import matplotlib.pyplot as plt
>>> label = MGS.get_label()
>>> for i in range(Nmgs):
...   ids = np.where(label==i)[0]
...   plt.scatter(data[:,0][ids],data[:,1][ids],c='C%s'%i)
...
>>> plt.show()

Another case

• The MGS uses Voronoi tessellation when it estimates density. But, sometimes it cannot calculate Voronoi cell because of several reasons. So, we also provide local density method to estimate density. The following example shows how to use 'local density' method.

>>> import Mulguisin
>>> import numpy as np
>>> data = np.random.random((10,2))
>>> Rcut = 0.1
>>> radius = 0.5 # This is the radius for local density method
>>> mgs_init = Mulguisin.mulguisin_type('local')
>>> MGS = mgs_init(Rcut,data[:,0],data[:,1],radius=radius)
>>> Nmgs, imgs, clg, clm, cng = MGS.get_mgs()
>>> Nmgs
9

More examples are in the test. We provide jupyter notebook to test MGS code.