Handling Speed and Pace with activatr

activatr has utility functions to easily annotate a GPX DF with information about the speed (and corresponding pace) of an activity.

First, let’s parse our sample GPX DF:

library(activatr)

# Get the running_example.gpx file included with this package.
filename <- system.file(
  "extdata",
  "running_example.gpx.gz",
  package = "activatr"
)

df <- parse_gpx(filename)

lat	lon	ele	time
37.80405	-122.4267	17.0	2018-11-03 14:24:45
37.80406	-122.4267	16.8	2018-11-03 14:24:46
37.80408	-122.4266	17.0	2018-11-03 14:24:48
37.80409	-122.4266	17.0	2018-11-03 14:24:49
37.80409	-122.4265	17.2	2018-11-03 14:24:50

This has information about location and time, but it’s often useful to visualize the “mile pace” over the course of an activity, especially for runs.

The first thing to add to the table is the speed. mutate_with_speed is a wrapper around dyplr::mutate that will add a speed column to the data frame:

df <- mutate_with_speed(df)

lat	lon	ele	time	speed
37.80405	-122.4267	17.0	2018-11-03 14:24:45	NA
37.80406	-122.4267	16.8	2018-11-03 14:24:46	1.872248
37.80408	-122.4266	17.0	2018-11-03 14:24:48	2.253735
37.80409	-122.4266	17.0	2018-11-03 14:24:49	2.897343
37.80409	-122.4265	17.2	2018-11-03 14:24:50	3.052919

However, speed (in meters per second) isn’t the most useful metric for runs; mile pace (in minutes and seconds) is a more common measurement. We can convert speed to pace using speed_to_mile_pace:

df$pace <- speed_to_mile_pace(df$speed)

lat	lon	ele	time	speed	pace
37.80405	-122.4267	17.0	2018-11-03 14:24:45	NA	NA
37.80406	-122.4267	16.8	2018-11-03 14:24:46	1.872248	859.576124817367s (~14.33 minutes)
37.80408	-122.4266	17.0	2018-11-03 14:24:48	2.253735	714.076761675982s (~11.9 minutes)
37.80409	-122.4266	17.0	2018-11-03 14:24:49	2.897343	555.453786757008s (~9.26 minutes)
37.80409	-122.4265	17.2	2018-11-03 14:24:50	3.052919	527.147986594496s (~8.79 minutes)

Finally, while pace here is a lubridate duration object, it’s most easily understood in a “MM:SS” format: pace_formatter is provided for that: contrast the readability of the two pace graphs below, the first plotting pace unmodified, the second reversing the axis and adding the formatter.

library(ggplot2)
library(dplyr)
library(lubridate)
ggplot(filter(df, as.numeric(pace) < 1200)) +
  geom_line(aes(x = time, y = as.numeric(pace)), color = "blue")
ggplot(filter(df, as.numeric(pace) < 1200)) +
  geom_line(aes(x = time, y = as.numeric(pace)), color = "blue") +
  scale_y_reverse(label = pace_formatter)

Note that the speed and pace are calculated at every point, so they will often be somewhat noisy. Computing the rolling mean of the speed to smooth out the graph would be numerically inaccurate, so mutate_with_speed provides helper arguments to allow the speed to be computed over larger windows.

In particular, it has lag and lead (which default to 1 and 0 respectively), which determines the “start” and “end” points used for the speed computation. So if we wanted each point’s speed to be determined using the points ten ahead and ten behind and plot that, we could do:

df <- mutate_with_speed(df, lead = 10, lag = 10)
df$pace <- speed_to_mile_pace(df$speed)
ggplot(filter(df, as.numeric(pace) < 1200)) +
  geom_line(aes(x = time, y = as.numeric(pace)), color = "blue") +
  scale_y_reverse(label = pace_formatter) +
  xlab("Time") +
  ylab("Pace")

Daniel Schafer