Linear regression for data with measurement errors and intrinsic scatter
BCES
Python module for performing robust linear regression on (X,Y) data points where both X and Y have measurement errors.
The fitting method is the bivariate correlated errors and intrinsic scatter (BCES) and follows the description given in Akritas & Bershady. 1996, ApJ. Some of the advantages of BCES regression compared to ordinary least squares fitting (quoted from Akritas & Bershady 1996):
- it allows for measurement errors on both variables
- it permits the measurement errors for the two variables to be dependent
- it permits the magnitudes of the measurement errors to depend on the measurements
- other “symmetric” lines such as the bisector and the orthogonal regression can be constructed.
In order to understand how to perform and interpret the regression results, please read the paper.
Installation
Using pip: