- Please refer to my previous post: Learnings from R, ggmap and ggplot2: I() the as is/inhibit/insulate function
- Hilarious that I didn’t remember my lesson from the above post from March 2016!
- Here’s the diff from sqlite-logo-text-hourly-ff-question-barcode.R so future me can remember better to use the ‘as is’ function aka
I()to Inhibit (“I” is for Inhibit remember :-) !) the conversion of the
os_coloursvector to factors (see What does the capital letter “I” in R linear regression formula mean?) and literally use this vector as a vector of colour names instead of requiring an additional line with:
print(image_df) p <- ggplot(image_df, aes(x, y)) + geom_image(aes(image = icon)) + - geom_label_repel(aes(label = question, colour = os_colours), + geom_label_repel(aes(label = question, colour = I(os_colours)), vjust = "inward", hjust = "inward", parse = TRUE, nudge_y = 2.5, nudge_x = -8, size = 3) + - scale_colour_identity() + theme_void() + expand_limits(y = c(0, 60), x = c(0, 60)) png_filename <- sprintf("logo-text-hour-%2.2d-%s.png", hour, base_name)