Interactive Plots with dygraphs

dygraphs allows us to make interactive plots in webpages.

Last Updated 2020-09-03

On previous posts, I showed how to make plots with the base and ggplot2 libraries. Today, I will make interactive plots with dygraphs. This package has a terrific documentation website here.

First, let’s load the data and the functions:

# import funs
source("./myfuns.R")

# import data
y <- import_gdp("../../data/PIB_BASE_2000.csv")

# visualization
head(y)

1980:Q1 1980:Q2 1980:Q3 1980:Q4 1981:Q1 1981:Q2 
   70.6    76.3    75.5    73.2    70.5    74.3 

Now, we load dygraphs and try to make a plot:

# libraries
library(dygraphs)

# plot
dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros")

OH! That worked out of the box, with similar syntax as base::plot(). Take that ggplot!

Ok, let’s try some options. First, let’s label the series as “GDP”.

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE)

Yes, we have to use the pipe %>%, so not EVERYTHING is similar to the base::plot(). As I said about ggplot, this is a fancy package. Let’s have go at the range selector option.

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE) %>%
dyRangeSelector(height = 20)

COOL! Now, let’s try to make lines. dygraphs uses dyEvent() for vertical lines and dyLimit() for horizontal lines.

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE) %>%
dyEvent("1983-03-30", "LABEL",  labelLoc = "top") %>%
dyEvent("2010-03-30", "LABEL",  labelLoc = "bottom") %>%
dyLimit( as.numeric(100), "LABEL", labelLoc = "right", color="red" )

So far, so good, now, let’s make those rectangles. dygraphs uses dyShading() to make those. Before, we have to load the recession limit dates.

# load recs
recs  <- qtr2date( readRDS("../../data/gdp-recessions.rds") )

# visualization
head(recs)

[1] "1981-03-31" "1983-03-31" "1987-09-30" "1988-09-30" "1989-09-30"
[6] "1992-03-31"

# tranform recs to data frame
recs.mat <- t( matrix( recs, nrow=2 ) ); colnames(recs.mat) <- c("start", "end"); recs.mat

     start        end         
[1,] "1981-03-31" "1983-03-31"
[2,] "1987-09-30" "1988-09-30"
[3,] "1989-09-30" "1992-03-31"
[4,] "1995-06-30" "1995-09-30"
[5,] "1998-03-31" "1999-03-31"
[6,] "2001-06-30" "2001-12-31"
[7,] "2003-03-31" "2003-06-30"
[8,] "2008-12-31" "2009-03-31"

# plot
dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE) %>%
dyShading( recs.mat[1, 1], recs.mat[1, 2] ) %>%
dyShading( recs.mat[2, 1], recs.mat[2, 2] ) %>%
dyShading( recs.mat[3, 1], recs.mat[3, 2] ) %>%
dyShading( recs.mat[4, 1], recs.mat[4, 2] ) %>%
dyShading( recs.mat[5, 1], recs.mat[5, 2] ) %>%
dyShading( recs.mat[6, 1], recs.mat[6, 2] ) %>%
dyShading( recs.mat[7, 1], recs.mat[7, 2] ) %>%
dyShading( recs.mat[8, 1], recs.mat[8, 2] )

NICE! But it seems a lot of dyShading lines, can we make it like in ggplot?

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE) %>%
dyShading( recs.mat)

Error in asISO8601Time(to): argument "to" is missing, with no default

OH! Do I have to provide two arguments?

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE) %>%
dyShading( recs.mat, recs.mat )

NOPE! Should I define the dimensions?

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE) %>%
dyShading( recs.mat[, 1], recs.mat[, 2] )

No, again. Oh well, should I try a for loop?

nrecs <- length(recs)/2; nrecs

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
dyAxis("x", drawGrid = FALSE) %>%
for(i in 1:nrecs) dyShading( recs.mat[i, 1], recs.mat[i, 2] ) %>%

Error: <text>:7:0: unexpected end of input
5: dyAxis("x", drawGrid = FALSE) %>%
6: for(i in 1:nrecs) dyShading( recs.mat[i, 1], recs.mat[i, 2] ) %>%
  ^

$%&*# !!!

OK, calm down, maybe it’s the pipe in the end, let’s change the order of the lines:

dygraph(y, main = "Cronologia Trimestral dos Ciclos de Negócios Brasileiros") %>%
dySeries(label = "GDP") %>%
for(i in 1:nrecs) dyShading( recs.mat[i, 1], recs.mat[i, 2] ) %>%
dyAxis("x", drawGrid = FALSE)

Error in function_list[[k]](value): object 'i' not found

Well, it wasn’t meant to be. But hey, the dyShading feature works, I just don’t know how to automate them. If you had more patience than me, give me some feedback on the comments.

Thank you for reading.