UPCOMING WEBINAR: The complete guide to analyzing NPS Register now >>

Articles in this section

See more

How to Create a Heatmap using R and HTML

  • Updated
Follow

You can quickly add a heatmap to your report via our Visualization selector, see How to Create a Heatmap. But for users who want to create a heatmap using special calculations or customizations, you can also create a heatmap in a Custom Calculation using R and HTML code. 

This article describes how to go from a grid table...

mceclip0.png

...to a customized heatmap:

mceclip1.png

Requirements

Method

1. From the toolbar, select the Calculation Screenshot 2024-04-22 133830.png > Custom Code and click on the page to create a custom R calculation.

2. The Code Editor should pop up at the top to edit the R code. If not, you can access this by clicking the object inspector > Data > R Code > Edit Code button.

3. Paste the code below into the editor. It creates a function called MyHeatmap. The last line uses that function on a table that you provide. Click inside the () on the last line and then click on the table you want to use to automatically insert its reference name into the code:

#create a function that will create a heatmap on a table you provide
MyHeatmap <- function(x)
{ 
# Lookups for coloring cells and fonts
require(RColorBrewer)
cell.colors = colorRampPalette(brewer.pal(9,"Blues"))(101)
font.colors = c(rep("blue", 60), rep("white", 41))

# Scaling the data to be on [1, 2, ..., 101], for the lookups
min.x <- min(x)
max.x <- max(x)
scaled.x = round((x - min(x)) / (max(x) - min(x)) * 100, 0) + 1

# Writing the cells styles
n.rows = nrow(x)
n.columns = ncol(x)
rows = rep(1:n.rows, rep(n.columns, n.rows))
columns = rep(1:n.columns, n.rows)
x.lookups = as.numeric(t(scaled.x))
cells.styles = paste0('td.cell', rows, columns, ' {background-color: ', cell.colors[x.lookups], '; color: ', font.colors[x.lookups] ,';}')
cell.styles = paste0(cells.styles, collapse = "\n")

# Creating the table
columns.headers = paste0('<th>', c("", dimnames(x)[[2]]) ,'</th>')
tble = paste0(columns.headers, collapse = "\n")
row.headers = paste0('<th>', c(dimnames(x)[[1]]) ,'</th>')
print('row.headers')
print(row.headers)
for (row in 1:n.rows)
{
row.cells = paste0('<td class = "border cell', row, 1:n.columns, '">', x[row,],'</td>')
print('row.cells')
print(row.cells)
tble = paste0(tble, '<tr>', row.headers[row], paste(row.cells, collapse = '\n'), '</tr>')
}
# Assembling the HTML
html = paste0('<!DOCTYPE html>
<html>
<head>
<style>
table, th, td {border-collapse: collapse; }
th, td {padding: 5px; text-align: center; font-family: arial; font-size: 8pt}
td.border {border: 1px solid grey; }',' td.cell {background-color: #F7FBFF; color: blue;}',
cell.styles, '
</style>
</head>
<body>
<table style="width:100%">', tble, '
</table>
</body>
</html>')
rhtmlMetro::Box(text = html, text.as.html = TRUE)
}

#click into () and click on the table to insert its name into the code and run the heatmap on it
MyHeatmap()

 

 

Next

How to Create a Table with Conditional Formatting using R and HTML

Related articles