Sick With Mono

The Data: Heart Rate and Temperature Measurements While Sick

Somehow I managed to get mono during a global pandemic at the age of 28. While I was sick I kept track of my temperature just using an oral thermometer, noting the time and temperature. My heart rate data comes from my Fitbit and represents my resting heart rate for the day. Here’s a glimpse of the data (first 10 rows):

## [1] "temps"

## # A tibble: 10 × 3
##    Date       Time                 Temp
##    <date>     <dttm>              <dbl>
##  1 2021-05-11 1899-12-31 19:48:00 101. 
##  2 2021-05-11 1899-12-31 20:41:00 101. 
##  3 2021-05-12 1899-12-31 08:13:00  98.2
##  4 2021-05-12 1899-12-31 13:38:00 100. 
##  5 2021-05-12 1899-12-31 16:42:00  98.6
##  6 2021-05-12 1899-12-31 18:29:00 100. 
##  7 2021-05-12 1899-12-31 19:22:00 101. 
##  8 2021-05-12 1899-12-31 19:39:00 101. 
##  9 2021-05-12 1899-12-31 20:09:00 101. 
## 10 2021-05-12 1899-12-31 20:37:00 101.

## [1] "hr"

## # A tibble: 10 × 2
##    Date          HR
##    <date>     <dbl>
##  1 2021-05-01    76
##  2 2021-05-02    74
##  3 2021-05-03    74
##  4 2021-05-04    75
##  5 2021-05-05    75
##  6 2021-05-06    78
##  7 2021-05-07    78
##  8 2021-05-08    77
##  9 2021-05-09    79
## 10 2021-05-10    81

Why does the time column have dates from the year 1899? Probably an Excel thing. Doesn’t matter for my purposes, I’ll only use the HH:MM:SS data from that column.

The Visualization: How does my heart rate and temperature change over time?

The Details: How the Plot was Made

For this visualization we need to load the following packages: tidyverse, lubridate, patchwork, extrafont, showtext. lubridate will help us handle working with datetime. (FAQ: Isn’t lubridate a part of tidyverse? Yes but. Yes, it gets installed when you install tidyverse. BUT it isn’t one of the “core” tidyverse packages that gets loaded when you load tidyverse so you need to load it separately.) patchwork will be used to combine the two plots. extrafont and showtext are for the font used in the plot title.

I had to fiddle with chunk options a bit to get the font to display in rmarkdown. Here’s what I did ( source):

• Load extrafont and showtext.
• Import the desired font from Google Fonts using font_add_google("Merriweather").
• Run showtext_auto(enable = TRUE) to automatically render the fonts.
• In the setup chunk include fig_retina = 1 so that the fonts in the HTML output render at the correct size.
• In the chunk(s) where you are displaying the visualization include fig.showtext = TRUE in the chunk options.

Here is the code for the two plots separately:

temp_plot <- temps %>% ggplot(aes(x = Time, y = Temp, color = Date, group = Date)) +
  geom_hline(yintercept = 100, lty = "dashed", color = "red") +
  geom_hline(yintercept = 98.6, lty = "dashed", color = "black") +
  geom_point(size = 3) +
  coord_cartesian(ylim = c(98,102)) +
  scale_color_viridis_c(option = "A", trans = "date", end = 0.95) +
  scale_x_datetime(date_labels = "%I %p", date_breaks = "2 hours") +
  labs(x = "Time of Day", y = "Temperature (F)") +
  theme_classic() +
  theme(plot.background = element_rect(fill = "grey95", color = "grey95"),
        panel.background = element_rect(fill = "grey95"),
        legend.background = element_rect(fill = "grey95"))

hr_plot <- hr %>%
  ggplot(aes(x = Date, y = HR, color = Date)) +
  geom_point(size = 3) +
  geom_line(size = 1) +
  scale_color_viridis_c(option = "A", trans = "date", end = 0.95) +
  scale_x_date(date_labels = "%b \n %d", date_breaks = "3 days") +
  labs(x = "Date", y = "Heart Rate (bpm)") +
  theme_classic() +
  theme(plot.background = element_rect(fill = "grey95", color = "grey95"),
        panel.background = element_rect(fill = "grey95"),
        legend.background = element_rect(fill = "grey95"),
        legend.position = "none")

( Here is where I learned about datetime formats.)

Next, I used patchwork to arrange the plots and add/format the title. With patchwork you can create a layout just by adding and/or dividing the plots. I also need to point out here that my code chunk options are the following for this chunk: {r fig.height=5, fig.width=10.5, fig.showtext = TRUE}.

hr_plot + temp_plot + plot_layout(widths = c(3.5,7)) + 
  plot_annotation(title = "I Got Mono and Graphed It", 
                  subtitle = "My heart rate and body temperature during the two weeks I was out sick.",
                  theme = theme(plot.title = element_text(family = "Merriweather", size = 20, hjust = 0.07),
                                plot.subtitle = element_text(family = "Merriweather", hjust = 0.095),
                                plot.background = element_rect(fill = "grey95")))

Why the weird hjust values? The default justification is left, all the way at the border of the plot, but I wanted the text to start closer to the y-axis so I scooted the text over a bit using hjust.