Comparison of canonical correspondence analysis and redundant analysis based on constrained ordination methods: an application on A cardiological data set

The events in nature arise with the effect of a large number of variables or factors that are related to each other, either individually or together. Redundacy Analysis (RDA) and Canonical Correspondence Analysis (CCA) are the methods of analysis used to determine these relationships. Studies on constrained canonical analysis methods, also known as Direct Gradient Analyzes (DGA), have generally been carried out by ecologists and there are limited number of studies in the field of health. Therefore, it is aimed to examine the CCA and RDA in this study. For this purpose, a data set including variables at different scale levels obtained from cardiyology field was used to compare RDA and CCA methods performance. In addition, the coefficients of determination (R2) and MAPE values were calculated as performance measures of the constrained ordination methods used in the study. According to the findings, it was found that the constrained ordination methods (CCA and RDA), which analyze the relationship structures between the variables found in different scale types, adequately explain the variation. In addition, it is emphasized that both methods have a good classification with a low MAPE value (less than 10%) and produce good ordination diagram thus, they can be used in the health field. Key words: Canonical form, scaling, biplot, dimention reduction, MAPE, angina