Plot_ly Scatterplot connecting lines how do most elegant? - r

Question

I'm asking myself how to solve the following problem the most elegant. My data encompasses of some actual values and some proposed values. Right now I have data that looks like the reproducible example below:
library(plotly)
library(dplyr)
test_dt <- data.frame(Age=1:5, Key=c("Actuals", "Actuals", "Actuals", "Other", "Other") , Value=rnorm(5))
plot_ly(data = (test_dt %>% group_by(., Key) %>% arrange(desc(Age))),
x = ~Age,
y = ~Value,
type = 'scatter',
mode = 'lines',
color = ~Key,
linetype = ~Key
) %>% layout(
yaxis = list(
title = "SD"),
margin = list(top=100, b=50)
)
The output of this code looks like this:
how plot a dashed line where i drew the red arrow?
My solution so far is that I access the last value of my actuals and insert this value as a new row for my "other" line. But I don't think that's very elegant and sometimes, if no other values exist which can happen in my data depending on the inputs then I have a legend plotted for my "other" line without actually having one.
act_age_max <- filter(test_dt, Key=="Actuals") %>% .[["Age"]] %>% max
propval_names <- filter(test_dt, Key!="Actuals") %>% .[["Key"]]
last_actual <- filter(test_dt, Age==act_age_max, Key=="Actuals") %>% .[["Value"]]
acts_year <- filter(test_dt, Age==act_age_max, Key=="Actuals") %>% .[["Year"]]
append_dt <- data.frame(Age=act_age_max, Key=propval_names, Value=last_actual)
plot_data <- rbind(test_dt, append_dt)
plot_ly(data = (plot_data %>% group_by(., Key) %>% arrange(desc(Age))),
x = ~Age,
y = ~Value,
type = 'scatter',
mode = 'lines',
color = ~Key,
linetype = ~Key
) %>% layout(
yaxis = list(
title = "SD"),
margin = list(top=100, b=50)
)