Astronomy and Astrophysics – Astronomy
Scientific paper
May 1998
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=1998aas...192.1102p&link_type=abstract
American Astronomical Society, 192nd AAS Meeting, #11.02; Bulletin of the American Astronomical Society, Vol. 30, p.831
Astronomy and Astrophysics
Astronomy
Scientific paper
Most astronomical data can be approximated as what one may call one-sided truncated data. A commonly occuring example of this is flux (or magnitude) limited data which includes all sources with flux f > fmin (or magnitude m < mmax). However, this is often a poor approximation and, in addition, there is a considerable amount of data that suffers from multiple truncations. A simple example of this is a sample with both upper and lower flux limits, fmax > f > fmin, such as recent quasar surveys. There have not been any general (e.g. non-parametric) methods for treating such data. Recently we have developed new non-parametric statistical methods that test for correlation and determine the underlying density functions of arbitrarily truncated data. These methods take into account the most general case of truncation, in which each data point has a different truncation region. Consequently they allow us to combine different samples with varied selection criteria. We describe these methods and apply them to combined samples of quasars. We show how this test of correlation between luminosity and redshift can be used to determine the quasar luminosity evolution. After removing this correlation parametrically we demonstrate how the method is used to determine the luminosity function, the co-moving density evolution and luminosity density as a function of redshift.
Efron Bradley
Maloney Alexander
Petrosian Vahe
No associations
LandOfFree
The Determination of Correlation and Distribution Functions for Multiply Truncated Data does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with The Determination of Correlation and Distribution Functions for Multiply Truncated Data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and The Determination of Correlation and Distribution Functions for Multiply Truncated Data will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1364141