Reputation: 4243
I am following the tutorial from:
https://rstudio.github.io/dygraphs/
They show a prediction at the bottom and I want to add it to the original time series graph. So below is the code and I am not sure how to combine the two:
install.packages("dygraphs")
library(dygraphs)
library(forecast)
library(dplyr)
dygraph(ldeaths)
Then I made a prediction:
hw <- HoltWinters(ldeaths)
predicted <- predict(hw, n.ahead = 72, prediction.interval = TRUE)
dygraph(predicted, main = "Predicted Lung Deaths (UK)") %>%
dyAxis("x", drawGrid = FALSE) %>%
dySeries(c("lwr", "fit", "upr"), label = "Deaths") %>%
dyOptions(colors = RColorBrewer::brewer.pal(3, "Set1"))
How do I combine the two into one graph?
Upvotes: 2
Views: 870
Reputation: 206546
Seem this library requires you to combine the data and then add separate series "layers" to the plot. With your example you could do this:
# combine the time series actual and forcasted values
combined <- cbind(predicted, actual=ldeaths)
# plot the different values as different series
dygraph(combined , main = "Predicted Lung Deaths (UK)") %>%
dyAxis("x", drawGrid = FALSE) %>%
dySeries("actual", label = "actual") %>%
dySeries(paste0("predicted.", c("lwr", "fit", "upr")), label = "Predicted") %>%
dyOptions(colors = RColorBrewer::brewer.pal(3, "Set1"))
Upvotes: 3