Model based prediction of Weight

Niraj Yadav
04/30/2017

This application predict Weight based on Height. Its based on a model created from data set which contains “Average Heights and Weights for American Women”

Data Set Name: women

Details

This app is trying to build a model to correlate height and weight of subject. It builds that model based on “women” dataset. Let's look at this dataset

data(women)
head(women)
  height weight
1     58    115
2     59    117
3     60    120
4     61    123
5     62    126
6     63    129

Prediction Model

Let's build this model

model <- lm(weight ~ height,data = women)

Let's take a sample input and predict weight

input <- 67

Predicted Weight

predictedWt <- predict(model, newdata = data.frame(height = input))
predictedWt
       1 
143.6333 

UI Details

This app takes Height as Input from Sliderbox and tries to predict weight

modelpred <- reactive({
     heightInput <- input$height
     predict(model, newdata = data.frame(height = heightInput))
   })

It renders output as text

output$pred <- renderText({
     modelpred()
   })

Plot Results

It also Plots the result and model in a chart

 plot(women$height, women$weight, xlab = "Height", 
          ylab = "Weight", bty = "n", pch = 16,
          xlim = c(55, 75), ylim = c(100, 180))
      if(input$showModel){
        abline(model, col = "red", lwd = 2)
      }
     points(heightInput, modelpred(), col = "blue", pch = 16, cex = 2)
     abline(h=modelpred(),col=4,lty=3) 
     abline(v=heightInput,col=4,lty=3)

   })