Developing Data Products Course Project: Shiny App Reproducible Presentation

There are many factors that can affect the mileage (miles per gallon/mpg) of an automobile. Using the mtcars dataset in the datasetspackage, a linear model was fit using the following variables:

• am : Transmission (automatic, manual)
• cyl : Number of cylinder (4, 6, or 8)
• disp : Displacement (cu.in)
• wt : Weight (lbs)

The shiny app created will predict the MPG (US gal) of an automobile based on the user's input of value of each of the above variables. Try it out!

The data was loaded from the datasets package and then formatted in order to better understand the variables.

#load data
data(mtcars)

#reformat variables
mtcars2 <- within(mtcars, {
    am <- factor(am, labels = c("automatic", "manual")) #transmission
    cyl  <- ordered(cyl) #number of cylinders
    wt <- 1000 * wt #weight in lbs
    })

The linear model used to predict MPG was the final model from a previous project. In summary, the variables chosen for the model were selected through ANOVA table.

fit_final <- lm(mpg ~ disp + wt + cyl + am, data = mtcars2)

You can view the project here.

This is the summary of the model.

                Estimate  Std. Error    t value     Pr(>|t|)
(Intercept) 30.275004533 3.290561817  9.2005579 1.166979e-09
disp         0.001632161 0.013756939  0.1186427 9.064703e-01
wt          -0.003249176 0.001249098 -2.6012170 1.512685e-02
cyl.L       -4.467787517 1.872176610 -2.3864135 2.458094e-02
cyl.Q        0.935361845 1.052357698  0.8888250 3.822469e-01
ammanual     0.141212000 1.326751152  0.1064344 9.160547e-01