Multivariate Statistics on HI and IRAS Images
Abstract
A procedure based on multivariate statistical methods have been worked out, making possible the simultaneous study of multiwavelength data with large number of variables. Testing the method a data set of HI (21cm) spectra and IRAS surface brightnesses from 891 positions of a Cep-Cas region have been analysed. The 891 HI spectra were replaced by 891times 3 factor coefficients: the new coordinates in the space of the first 3 factors (as orthogonal axes) resulted by a principal component analysis (PCA). A cluster analysis have been performed in the 7 dimensional parameter space of 3 factor coefficients and 4 IRAS fluxes. Results have been compared to the previous knowledge on the area.
- Publication:
-
Astronomical Data Analysis Software and Systems I
- Pub Date:
- 1992
- Bibcode:
- 1992ASPC...25..251T