In this paper, comparison results of parametric methodologies of change points, applied to maximum temperature records from the municipality of Tlaxco, Tlaxcala, México, are presented. Methodologies considered are likelihood ratio test, score test, and binary segmentation (BS), pruned exact linear time (PELT), and segment neighborhood (SN). In order to compare such methodologies, a quality analysis of the data was performed; in addition, lost data were estimated with linear regression, and finally, SARIMA models were adjusted.

In this work, an analysis of change points is made with the Bayes factor, thresholds, and cumulative sum (CUSUM) statistics methods. For the analysis of change points with the Bayes factor, Poisson data were simulated; the threshold method was worked with a regression and data of the National Institute of Statistics, Geography and Informatics (INEGI) of Mexico and coronavirus were used for the CUSUM.