Visualizing the Output of datsteps()
To get a general impression of the data set and the possibilities of visualization, we explore and recommend different plot methods below. The simplest and fastest way to plot the output with base R is the plot() function in combination with the density() function:
plot(density(inscr_steps$DAT_step))
Using ggplot2
A somewhat crowded overview containing the Locations of all Inscriptions can easily be achieved when using ggplot2 and its geom_density()-method, which is based on the same procedure as the density()-function used above. A result may look like this:
ggplot(data = inscr_steps, aes(x = DAT_step, fill = variable,
weight = weight)) +
geom_density(alpha = 0.3)
Note that the output without the weights calculated by datsteps() is very different:
ggplot(data = inscr_steps, aes(x = DAT_step, fill = variable)) +
geom_density(alpha = 0.3)
The weights calculated by datplot (internally using the get.weights()-function) are a way of assigning an importance to the objects in question. They relate to the range an object is dated to (more detailed information on the process is available in (Weissova and Steinmann, n.d.)). Therefore, any objects dated to large time spans as for example “Roman Imperial Period” ranging from 31 BC to 395 AD in our data set, are contributing to the curve significantly less than any object dated to one exact year. At times it can be very useful to look at both outputs and discuss them separately. We will look at a case study of a single city later on to clarify this.
Using ggplot2 and ggridges
Since the graph is very crowded in this layout, we actually recommend using geom_density_ridges from the ggridges-package. Since the package has no built in support for assigning weight, we have to use stat = "density", which will reference the density()-function to get the appropriate calculations:
ggplot(data = inscr_steps,
aes(x = DAT_step,
y = fct_rev(as_factor(variable)),
fill = variable,
weight = weight)) +
geom_density_ridges(aes(height = ..density..),
stat = "density", alpha = 0.9) +
scale_fill_discrete(guide = FALSE)
Styling will help to make the plots more readable:
bluegreen <- colorRampPalette(c("#8dae25", "#17365c"))
ggplot(data = inscr_steps,
aes(x = DAT_step,
y = fct_rev(as_factor(variable)),
fill = variable,
weight = weight)) +
geom_density_ridges(aes(height = ..density..), stat = "density", alpha = 0.9) +
scale_x_continuous(breaks = seq(from = -800, to = 800, by = 100),
limits = c(-800,800), name = "") +
geom_vline(xintercept = 0, alpha = 0.5, lwd = 1) +
theme(axis.text.x = element_text(angle = 90, vjust = 0.5),
panel.background = element_blank(),
panel.grid.major.x = element_line(linetype = "dashed",
color = "gray30"),
panel.grid.minor.x = element_line(linetype = "dotted",
color = "gray80")) +
scale_fill_manual(guide=FALSE,
values = bluegreen(length(unique(inscr_steps$variable)))) +
labs(title = "Epigraphic Evidence from Bithynia",
subtitle = "Spatio-temporal distribution",
y = "administrative centres",
caption = attributes(inscriptions)$source)
Using ggplot2 and facet_wrap()
Another option is to separate the variables with facet_wrap, which is not as condensed and does not support the option to scale the densities in the graph for maximum visibility, as geom_density_ridges() automatically does (see Documentation of ggridges). In our case, this leads to a worse visibility of smaller density curves:
ggplot(data = inscr_steps, aes(x = DAT_step,
fill = variable, weight = weight)) +
geom_density(alpha = 0.9) +
scale_fill_discrete(guide = FALSE) +
facet_wrap(variable ~ ., ncol = 1)
Styling also helps very much to improve the graphs readability:
ggplot(data = inscr_steps, aes(x = DAT_step,
fill = variable, weight = weight)) +
geom_density(alpha = 0.9) +
theme(panel.background = element_blank()) +
scale_fill_manual(guide=FALSE,
values = bluegreen(length(unique(inscr_steps$variable)))) +
scale_x_continuous(breaks = seq(from = -800, to = 800, by = 100),
limits = c(-800,800), name = "") +
facet_wrap(variable ~ ., ncol = 1) +
theme(strip.text.x = element_text(size=8),
strip.background = element_blank(),
axis.text.x = element_text(angle = 90, vjust = 0.5),
axis.text.y = element_blank(),
axis.ticks.y = element_blank(),
axis.title.y = element_blank(),
panel.background = element_blank(),
panel.grid.major.x = element_line(linetype = "dashed",
color = "gray30"),
panel.grid.minor.x = element_line(linetype = "dotted",
color = "gray80")) +
labs(title = "Epigraphic Evidence from Bithynia",
subtitle = "Spatio-temporal distribution",
caption = attributes(inscriptions)$source)
