Data based investigation on galaxy formation and evolution theory through statistical techniques

Abstract

The present work deals with a large data set of star-forming as well as quiescent galaxies at red shifts ranging from 0 to 4 (0$\le$z$\le$4). The present selected catalogues of objects have been compiled from the NEWFIRM Medium-Band Survey (NMBS). We have concatenated both the catalogues AEGIS and COSMOS, present in NMBS, one after another to form a master catalog. Several derivable parameters were present in the catalog, but we have limited our choice to the observable as well as the physical parameters of the galaxies. In this paper, we have studied the evolutionary pattern of the galaxies by dividing them into various groups according to red shift, and observing the distributional pattern of those groups. Later, the evolutionary pattern of the galaxies is investigated by examining the size-mass relationship corresponding to these groups and comparing the level of dependence of the parameters under study.

In order to perform a proper analysis of the data for the above-mentioned objectives, we have used statistical techniques like multiple testing, the Shapiro–Wilk test, independent component analysis, multivariate outlier detection, multivariate kernel density estimation, and kernel regularized least squares method. Two noble findings that have been observed in this work are (a) The galaxy parameters with high red shifts (z $>$ 3.5) follow a multivariate Gaussian distribution, helping us to infer that high red shift (z $>$ 3.5) galaxy parameters show more randomness compared to other galaxies with z$<$3.5. (b) There is a deviation from the linearity of the covariates in very high-redshift galaxies (z $>$ 3.5) for modeling log mass as a response variable. The same is also observed with the half-life radius as the response variable, although there exists a linear relationship between the mass and the effective radius of the galaxy. These observations may be treated as new findings of the present study.