Comparing plotly & ggplotly plot generation times

• Prerequisites
• Analysis
• Summary

Prerequisites

Throughout we will be using the packages: dplyr, tidyr, ggplot2, plotly and microbenchmark. The data in use is the birthdays dataset in the mosaicData package. This data sets contains the daily birth count in each state of the USA from 1969 – 1988. The packages can be installed in the usual way (remember you can install packages in parallel)

install.packages(c("mosaicData", "dplyr", 
                   "tidyr", "ggplot2", 
                   "plotly", "microbenchmark"))
library("mosaicData")
library("dplyr")
library("tidyr")
library("ggplot2")
library("plotly")
library("microbenchmark")

Analysis

Let’s load and take a look at the data.

data("Birthdays", package = "mosaicData")
head(Birthdays)
##   state year month day       date wday births
## 1    AK 1969     1   1 1969-01-01  Wed     14
## 2    AL 1969     1   1 1969-01-01  Wed    174
## 3    AR 1969     1   1 1969-01-01  Wed     78
## 4    AZ 1969     1   1 1969-01-01  Wed     84
## 5    CA 1969     1   1 1969-01-01  Wed    824
## 6    CO 1969     1   1 1969-01-01  Wed    100

First, we’ll create a very simple scatter graph of the mean births in every year.

meanb = Birthdays %>% 
    group_by(year) %>% 
    summarise(mean = mean(births))

Wrapping this as a ggplot object inside ggplotly() we obtain this…

ggplotly(ggplot(meanb) + 
  geom_point(aes(y = mean, x = year, colour = year)))

plot_ly(data = meanb, 
        y = ~mean, x = ~year, color = ~year, 
        type = "scatter")

Both graphs are, identical, bar styling, yes?

Now let’s use microbenchmark to see how their timings compare (for an overview on timing R functions, see our previous blog post).

time = microbenchmark::microbenchmark(
        ggplotly = ggplotly(ggplot(meanb) + 
                            geom_point(aes(y = mean, x = year, colour = year))),
        plotly = plot_ly(data = meanb, 
                         y = ~mean, x = ~year, 
                         color = ~year, type = "scatter"),
        times = 100, unit = "s")
time
## Unit: seconds
##      expr      min       lq     mean   median       uq     max neval cld
##  ggplotly 0.050139 0.052229 0.070750 0.054760 0.056785 1.56652   100   b
##    plotly 0.002475 0.002527 0.003017 0.002571 0.002674 0.03061   100  a
autoplot(time)

Now I thought nesting a ggplot object within ggplotly() would be slower than using plot_ly(), but I didn’t think it would be this slow. On average ggplotly() is approximately 23 times slower than plot_ly(). 23!

Let’s take it up a notch. There we were plotting only 20 points, what about if we plot over 20,000? Here we will plot the min, mean and max births on each day.

date = Birthdays %>% 
  group_by(date) %>% 
  summarise(mean = mean(births), min = min(births), max = max(births)) %>% 
  gather(birth_stat, value, -date)

Wrapping this a ggplot2 object inside ggplotly() we obtain this graph…

ggplotly(ggplot(date) +
    geom_point(aes(y = value, x = date, colour = birth_stat)))