Training For Jim Bridger Trail Run

Gabe Rudy
2015-07-26

Project Overview

map

To get in shape for a 10 mile mountain trail run called the Jim Bridger Trail Run on June 27, 2015, I started counting calaries using MyFitnessPal to drop to a target weight of 160.

With my iPhone (and Apple Watch starting early May) I had a trove of data colleted in HealthKit, and was able to export it using QS Access showing my training up to the race and after.

Measurements

map

I reached my goal, finished the race in 1:52, and placed 42 (out of ~120 men)

References:

Shiny App Construction

The Shiny App Used the following rPlots libraries:

xCharts
Morris
Richshaw

The left combo box used input resulted in a different component being rendered in a renderUI function. Allowing each plot to replace the main panel.

 output$plot <- renderUI({
        if(input$chart == "lbs"){
            showOutput("plot_lbs", "xcharts")
        }else if(input$chart == "cal"){
            showOutput("plot_cal", "morris")
        ...

Predictive Model

A simple predictive model was computed based on my training data to predict weight after N days. Here is the model computed for 100 input days:

by.day <- readRDS('jbtr/health_data_by_day.rds')

#Linear Model:
train <- by.day[1:which(by.day$Day == "2015-06-27")[1],]
train$n <- (1:nrow(train))
model <- lm(Weight.lb~n, data=train)
predict(model, newdata = data.frame(n = 100), interval="predict")

       fit      lwr     upr
1 175.9549 172.1478 179.762