It is possible to change the aesthetics of tokens based on the timestamp of the animation
For example, the number of days a ‘patient’ has been in the system
# Libraries ---------------------------------------------------------------
library(dplyr) ##pipes
library(tidyr) ##tidy data, partcularly the crossing() function
library(lubridate) ##date time manipulation
library(bupaR) ##buisness process analytics
library(processanimateR) ##animates process
# Create performance time flags ------------------------------------------------
my_flags <- data.frame(value = c(0,2,4,8,16)) %>%
mutate(day = days(value)) #convert numeric value into daysThis will change the colour of the token at 0, 2, 4, 8, and 16 days.
The crossing() function joins the cases of ‘patients’ to ‘my_flags’ and creates all possible combinations.
# Create timestamps of flags ----------------------------------------------
my_timeflags <- patients %>%
cases %>%
crossing(my_flags) %>% ##similar to a SQL outer join
mutate(time = start_timestamp + day) %>%
filter(time <= complete_timestamp) %>%
select("case" = patient,time,value) ##must be case, time, valueThe data for the token_scale() function must have the column headings ‘case, time, value’.
Without the domain = my_flags$value argument the flags follow alphabetic order (e.g. 0, 16, 2, 4, 8) rather than the numeric order we wants. See d3-legend for further information.
# Animate process ---------------------------------------------------------
patients %>%
animate_process(mode ="absolute",
jitter=10,
legend = "color",
mapping = token_aes(
color = token_scale(my_timeflags
, scale = "ordinal"
, domain = my_flags$value
, range = rev(RColorBrewer::brewer.pal(5,"Spectral"))
)))The colors can be modified through the range argument. In this case the scale is reversed with rev() to go from blue to red. See RColorBrewer::brewer.pal.info for all options:
Acknowledgement
Thanks to Dominic Rowney for this nice example of advanced processanimateR usage. The original example code can be found here.