PRIMARY AND VISUAL ANALYSIS OF ROWING PERFORMANCE DATA BY MEANS OF PYTHON USING PANDAS, MATPLOTLIB AND SEABORN LIBRARIES

Abstract

At present, computerization plays an important role in higher education, aimed at improving the educational process by shaping and improving its forms and content. In particular, computerization is applicable in the sphere of physical education as well. The article presents the use of Python programming language and its libraries Pandas, Matplotlib and Seaborn in the analysis of test data on academic rowing by students of different ages at Petro Mohyla Black Sea National University. Presented is the structure of the original data set for further analysis, which includes the characteristics measured before the test and recorded in the protocol, and the characteristics obtained by importing from the rowing machine monitor. The basic capabilities of the Pandas library to perform primary data analysis using the DataFrame structure are presented. Primary analysis can be used to identify students with the best or worst test scores, the average values of the walking distance, taking into account gender, age, nationality, etc. By means of visual analysis the dependencies between some quantitative characteristics were investigated. Graphs of pairwise dependencies between the distances covered for certain periods of time and graphs of the distribution of these values are presented. Correlation matrix for certain quantitative characteristics such as age, body weight, height, heart rate before and after the performance of the rowing test and the distances covered over a certain period of time is presented. A general scatter diagram of the traits of number of strokes and body weight of a student and a scatter diagram of these traits in the context of the categorical trait of gender are presented. Primary and visual data analysis can be used for more in-depth data analysis using machine learning and artificial intelligence methods and is the first step in creating a system of intelligent data analysis and prediction of sports performance, which can be used in the field of physical education and applied to monitor the results of different sports.