LinReg - 2020 Shiny Contest Submission


Authors: Pierre Bouchet
Working with Shiny < 1 year

Abstract: This app LinReg aims to make easier a linear regression with graphs and model checkings.

Full Description: This app LinReg aims to make easier a linear regression with much graphs and model checkings. For example, there are residuals vs fitted, summary, qq-plot, AIC and much more!

It is built to be more friendly with non-initiated persons than the usual procedure with R which can be tiresome.

There are 3 steps in this application :

  1. Data import. At the moment is reads only .CSV with column names. (I'm on it :wink: )
  2. Variable selection : The application asks you which column is the target variable and which variable(s) is (are) the explanatory variable(s).
  3. Linear regression : you have the summary table, the residuals vs fitted, qqplot, cook's distance, Box-Cox transformation, residuals ACF, residuals PACF, Shapiro-Wilk test, AIC, Durbin-Watson test, Variance Inflation Factor (only for 2+ expl. var.), Allen's PRESS, Leverages (or hatvalues) !
    I've made this app on my own and I'm new to RShiny so this is not perfect :smiley:
    It is built to make faster a quasi-complete linear regression analysis and above all to analyse residuals which are very important for the regression's validity.

If you want to test the application please don't hesitate to contact me before as I can give you a .CSV as a toy exemple :slight_smile:

Category: Consulting
Keywords: linear regression, data science tool
Shiny app:
RStudio Cloud: Posit Cloud

Full image: