Programming Correlation Criteria with Free CAS software, Computational Economics.(DOI: 10.1007/s10614-016-9604-1)
Programming Correlation Criteria with Free CAS software, Computational Economics, DOI: 10.1007/s10614-016-9604-1. (SCOPUS, EconLit, ISI Listed. Impacty factor: 0.521 and 5-year: 0.526). In common with K. Tsilika (2016).
Our contribution in this work is to set the directions for specialized econometric computations in a free computer algebra system, Xcas. We focus on the programming of a routine dedicated to correlation criteria for multiple regression models. We program several operations for detecting and evaluating collinearity by applying the diagnostic techniques of linear regression analysis. In order to illustrate the computational performance of our Xcas codes, we repeat and test most of the analysis carried out in widely used commercial software, along with some extra statistics. Xcas could constitute a supplemental tool in a collinear data study. Its use is proposed complementary to established econometric software or as substitute software.