apexcharter – Intro, Installation & Loading

apexchart.js is a modern JavaScript charting library to build interactive charts and visualizations with simple API. apexcharter is built as a htmlwidget (R Package) for apexchart.js and the API design is inspired by highcharter. apexcharter requires RStudio >= 1.2 to properly display charts.

Install the stable version from CRAN with:

install.packages("apexcharter")

Or install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("dreamRs/apexcharter")

Once successfully installed, apexcharter can be loaded using

library(apexcharter)

Simple Example

The main function of apexcharter is the apex() function whose first argument is data. Thus, enabling the support of pipe %>% operator. The second argument is mapping – aesthetics (x & y) and the third one is type of the chart – which takes multiple values like scatter, bar, line and much more.

Let’s take R’s in-built mtcars dataset and draw a simple bar chart.

library(apexcharter)
library(tidyverse)

mtcars %>% 
  count(cyl) %>% 
  apex(type = "bar",
       mapping = aes(x = "cyl", y = n))

Now, that’s a beautiful interctive chart. Let’s go ahead and see a few more examples of something bigger than a simple bar chart.

Building Interactive Heatmap / Correlation Plot

Let’s try to visualize a Heatmap (of Correlation Plot) of numeric columns of mtcars dataset. To do that, we’ve to first select the numeric columns which we’ll do with select_if(is.numeric) and then we’ve to build the correlation matrix which the base-R function cor() does it smoothly.

Now that we’ve got a matrix let’s convert it to a data.frame and for us to draw a heatmap – we need 3 things primarily:

• x-axis – categorical
• y-axis – categorical
• fill value – continuous

So, we’ll convert the rownames of the resultant dataframe to a column and then convert the wide format data into long format using pivot_longer().

At this point our data is in the desirable format for a heatmap. Simply for aesthetics improvement, let’s round off the correlation values.

Finally, we’ll use our apex() function with type = 'heatmap' that gives us a color-filled heatmap (that’s also interactive).

library(apexcharter)
library(tidyverse)

mtcars %>% 
  select_if(is.numeric) %>% 
  cor() %>% 
  as.data.frame() %>% 
  rownames_to_column("col") %>% 
  pivot_longer(cols = -col, names_to = "type") %>% 
  mutate(value = round(value,2)) %>% 
  apex(type = "heatmap",
       mapping = aes(x = col, y = type, fill = value)) 

Building Interactive Time-Series (Line) Graph

If there’s a plot where Interactive Charts are incredibly valuable, I think it’s Time-Series Graph where labelling on traditional (static) chart would sometimes make the chart clunky and less readable.

Let’s build an Interactive Time-series plot with the apexcharter library. As you can see below, all it takes is a dataframe with a column denoting the time field and another column with the actual value for that time.

library(apexcharter)

df <- data.frame(Y=as.matrix(EuStockMarkets), date=time(EuStockMarkets))

df %>% 
  apex(type = "line",
       mapping = aes(x = date, y = Y.DAX)) 

Summary

Thus, We learnt how to build interactive charts using apexcharter that follows a very minimal API similar to highcharter.

References

• apexcharter : dreamRs – Github
• apexcharts.com
• How to reshape a dataframe from wide to long or long to wide format using tidyr – pivot_longer() pivot_wider()