Automatizando a análise de regressão linear no RStudio : um aplicativo interativo em Shiny

Abstract

This work proposes the development of an interactive application in R language to automate linear regression analysis, a statistical technique widely used in various areas of knowledge. The application of linear regression analysis can be complex and requires knowledge in statistics and programming, which can be an obstacle for many users. In addition, it can be a time-consuming task, especially when dealing with large data sets and multiple variables. The proposed application aims to overcome these challenges, allowing users to perform linear regression analyses in an intuitive way and without the need for coding. With this application, users will be able to enter their own database, select the independent variables and the dependent variable, and perform a linear regression analysis. This makes the statistical technique more accessible and easy to use, saving users’ time and effort and allowing them to focus on interpreting the results instead of implementing the model. For the application test, a database of energy consumption from the company DAEWOO Steel Co., Ltd, a South Korean company (UCI, 2023), will be used. The response variable of interest is the industry’s energy consumption and the covariates are carbon dioxide, relative current power and days of the week. The built application allows the insertion of a database in Excel, presentation of the data in table format, descriptive analysis of the data, simple and multiple linear regression analysis, graphical analysis of the residuals and model selection.