Markov Chain Monte Carlo for Correlation Coefficient
The purpose of this program is to examine the correlation coefficient of a given linear fitting, with the consideration of data error in both coordinates.
Markov Chain Monte Carlo is iterated by WinBUGS, and manipulated by Matlab.
matbugs.m is an Matlab interface to WinBUGS, details.
Correlation.txt defines the correlation coefficient model for WinBUGS.
Regression.txt defines the linear regression model for WinBUGS.
EisoDeltaTCorr.m gives an example of computing correlation coefficient for Matlab.
EisoDeltaTReg.m gives an example of linear regression for Matlab.
Other files are the data and results.