A prediction formula for backpacking travel times
Problem Statement
Propose a practical backpacking travel time prediction model and fit it to past trip records.Summary
Good planning of backpack trips includes knowing how far you can go in one day. Over the years, I have used a prediction formula for the time it takes to hike a route, given the route's distance and total elevation gain. The formula works well enough for trails, but I don't know where it comes from. It was handed down to me, practically an oral tradition.More recently, I realized that this formula could be improved by taking into account the type of terrain, since a significant portion of my long Summer trips are cross country. About the same time, I acquired an accurate record of travel times for four major trips by carrying a GPS device with the tracking mode on.
Thirty-one days of backpacking travel time and route statistics
It is a textbook statistics problem to fit the backpacking travel time data to a new model. In this project, I estimated the coefficients for a model with three terms: (1) route miles on trail, (2) route miles cross country, and (3) total elevation gain. I also used the residuals to estimate the left-over uncertainty in the predictor.
Read the Details
- Documentation of the FredsEmpirical formula for estimating travel times from route distance .
- Export of Jupyter Notebook, plots of backpacking travel times and route statistics, code for multiple linear regression using Python StatsModels library.
- Documentation of the FredsEmpirical formula for estimating travel times from route distance.
- Paper, discussion of predictor variables, proposed prediction model and math equations.