Geom pandemics

[jmaltman]

created geom_pandemics.R (not ready)
updated geom_recessions.R with new URL
updated sysdata.rda
created pandemics.rda
created R script to create pandemics.rda

[jmaltman]

When trying to run geom_pandemics, this error occurs:
Error in geom_pandemics():
! Problem while setting up geom.
ℹ Error occurred in the 1st layer.
Caused by error in UseMethod():
! no applicable method for 'mutate' applied to an object of class "character"
This seems to be an issue from line 219 of the geom_pandemics code. Not sure why this is happening since the same code (minus name changes) exists in geom_recessions and that function works

[dlcomeaux]

@jmaltman heads up, I had a bit of time and was able to get the function to work. I did not address the geom_recessions data update link. Could you play around with that using the new format, ideally using the stable geoJSON link that they have now migrated to rather than the csv/xlsx approach we used before?

[dlcomeaux]

It would also be better practice for the code if we streamlined/reduced duplication between geom_recessions and geom_pandemics. This can certainly be done, and ideally we would have a consolidated documentation too - but woudl take some work and require some changes to arguments (the most important one of which would be changing update_recessions and update_pandemics to a consistent update parameter).

[dlcomeaux]

OK @jmaltman I would appreciate if you could see whether this works on your system, including in edge cases. I think it's a functional PR at this stage. When we are ready, we will need to increment the cmapplot version number and add a change log.

[jmaltman]

Testing this function out with the update piece of the function to insert my own pandemics data, I found two errors come up.

1. The labels for the made-up pandemics did not change to reflect the name of these imaginary pandemics. Instead, each of them kept the label "COVID-19 Pandemic" in the chart.
2. For an imaginary pandemic that was ongoing (where ongoing = TRUE in the data), if no end date was added to the ongoing pandemic, then it would not populate in the chart. If I gave it an imaginary date in the future, then it appeared, but if I left the end date as NA, then it did not appear when trying to plot it.

[dlcomeaux]

@jmaltman

Hmm, I am not able to replicate the bug for the labels. When I add a new pandemic table by copying the existing one and adding a new row, it shows up fine for me with the accurate label. Could you paste the code you used to yield this bug?

The ongoing one requires an end date by the way that the function currently works. You could add some code to do that automatically, if you wanted, but in the recessions table a user supplied one wouldn't work unless you put a far out date on it.

[jmaltman]

@dlcomeaux, I created my own data frame instead of adding to the existing pandemic table.

new_df <- data.frame(start_char = c("Jan 1980", "Apr 2009", "Mar 2015"),
end_char = c("Jun 1986", "Aug 2010", "Feb 2025"),
start_date = c("1/31/1980", "4/25/2009", "3/26/2015"),
end_date = c("6/15/1986", "8/10/2010", "2/1/2025"),
ongoing = c(FALSE, FALSE, TRUE),
name = c("Shoe", "Swine Flu", "Plushie")
)

new_df <-
new_df %>%
mutate(start_date = as.Date(start_date, format = "%m/%d/%Y"),
end_date = as.Date(end_date, format = "%m/%d/%Y"))

ggplot(df, mapping = aes(x = year_dec, y = value)) +
geom_pandemics(update = new_df) +
geom_line(aes(color = var)) +
scale_x_continuous("Year") +
theme_minimal()

[dlcomeaux]

@jmaltman OK this has identified a few things for me.

• One issue with this is the NA code - can you try what happens on your system if you give that one an end date? Do the labels work properly then?
• When I run that, it gives me an error that it may not be properly formatted, although it still outputs the geoms
• I think I know what is happening with the labels. The label is being extracted as a part of the geom before the update function is run, which means that it pulls from the built in labels.