for a research-project i want to plot a historical timeline with ruling-periods of ancient dynasties; seperated into male and female rulers.
I'm using R and the vistime-package (as suggested here). Generally it works, but the problem is that i can't figure out how to use dates BC; that means negative years.
A simple example of my code reads as follows:
# devtools::install_github("edgararuiz/gregorian")
# library(gregorian)
library(ggplot2)
library(vistime)
counter.chronology <- data.frame(
namen = c("Ptolemaios V.","Ptolemaios VI.","Kleopatra I.","Kleopatra II.","Kleopatra III."),
geschlecht = c("Männliche Regenten","Männliche Regenten","Weibliche Regentinnen","Weibliche Regentinnen","Weibliche Regentinnen"),
antritt = as.Date(c("0197-01-01","0180-01-01","0194-01-01","0175-01-01","0141-01-01")),
ende = as.Date(c("0180-01-01","0164-01-01","0176-01-01","0164-01-01","0130-01-01")),
stringsAsFactors = FALSE
)
vistime(counter.chronology, col.event = "namen", col.group = "geschlecht", col.start = "antritt", col.end = "ende")
This works fine and produces a plot like this:
plot with wrong (positive) dates
But if i change the date-format to negative years -- for example "-0175-01-01" -- the plotting doesn't work out and i receive the error-message:
Errore in charToDate(x) :
character string is not in a standard unambiguous format
I tried the gregorian-package and replaced as.Date with as_gregorian, but this seems imcompatible with vistime.
Does anyone know an easy solution for this problem? If negative dates are impossible, it would help to turn the plot around, in a way that it counts down an the x-axis from the highest to the lowest year.
A less important question, but also nice, if it's solved: It would be enough to use years as start and end dates. Months and days are unnecessary. But if i only enter, for example, "-0175 or the positive version "0175" the same error-message as above occurs. Therefore, i used 01-01 for month and day. It works anyway, because the timeline is not that detailed. To solve this would be nice, but it's not a must.
Thanks for all your replies and answers!
Best,
Flo
Edit after Allen response:
Another problem occured which you couldn't be aware of due to my short code-example. Sometimes there are overlapping ruling-periods. If i now enter another female reign which overlaps two of the other periods -- here, for example, "Arsinoe" -- , it looks awkward:
counter.chronology <- data.frame(
namen = c("Ptolemaios V.", "Ptolemaios VI.",
"Kleopatra I.","Arsinoe", "Kleopatra II.","Kleopatra III."),
geschlecht = rep(c("Männliche Regenten", "Weibliche Regentinnen"),
times = c(2, 4)),
antritt = -c(197, 180, 194, 200, 175, 141),
ende = -c(180, 164, 176, 150, 164, 130)
)
Is it possible to let ggplot place this overlapping bar automatically one stage above the others?
Found another way usin the timevis package. Works well so far, also with overlapping periods:
library(timevis)
counter.chronology <- data.frame(
content = c("Ptolemaios V.","Ptolemaios VI.","Kleopatra I.","Arsinoe","Kleopatra II.","Kleopatra III."),
start = c("-000197-01-01","-000180-01-01","-000194-01-01","-000190-01-01","-000175-01-01","-000141-01-01"),
end = c("-000180-01-01","-000164-01-01","-000160-01-01","-000145-01-01","-000164-01-01","-000130-01-01"),
group = c(1,1,2,2,2,2),
style = c("background-color: #f1d9a4; border-color: black;",
"background-color: #ceaf7a; border-color: black;",
"background-color: #ca9865; border-color: black;")
)
timevis(
counter.chronology,
groups = data.frame(
id = 1:2,
content = c("Regenten", "Regentinnen"),
width = 900
)
)
From a style-point i like the ggplot version better, but didn't found any solution for the problem with overlapping bars. Using gg_vistime would be my preferred solution -- since it combines the tools of vistime and ggplot. Unfortunately, like seen on the vistime-Github-page, there seems to be no practicable solution for using BC-dates, except for a complicated workaround: https://github.com/shosaco/vistime/issues/6
For now, I'm going with the timevis solution.
