How to Change the Tooltip in ggplotly to show SUM of a Column in a Bar Chart

Question

I have this bar chart rendered using ggplot and ggplotly()

It renders just the way I want, in that it displays the SUM of total crashes by Mode Type and Year.

But, I am struggling with getting the (hover) tooltip to display the SUM of the Total Crashes. Right now, it just shows "1".

What do I need to do to my code to configure the tool tip to show the total crashes as shown in the chart, and not a value of 1? Here is my code, and the bar chart it produces:

crashplot <- ggplot(totcrashes_by_year, aes(totcrashes_by_year$`Crash Year`, totcrashes_by_year$`Total Crashes`))
crashplot +geom_bar(stat = "identity", 
                        aes(fill=totcrashes_by_year$`Mode Type`), position = "stack") + xlab("Crash Year") + ylab("Total Crashes") + ggtitle("Total Crashes by Year and Mode Type") + labs(fill = "Mode Type")

EDIT: Reproducible Code Example Here is code for replicating this issue.

crashsample <- data.frame(year = sample(2007:2018, 40, replace = T),
             crashes = 1, 
             type = sample(c('Vehicle', 'Pedestrian','Bike','Motorcycle'), 5, replace = TRUE))

Create ggplot

crashplot <- ggplot(crashsample, aes(x =  year, y = crashes, fill = type)) +
  geom_bar(stat = "identity") + 
  labs(x = 'Crash Year', 
       y = 'Total Crashes', 
       title = "Total Crashes by Year and Mode Type", 
       fill = "Mode Type")
#call to ggplotly to render as a dynamic plotly chart
ggplotly(crashplot)

J. Ring · Accepted Answer · 2020-10-24T10:36:42.973

1

This code returns valid labels for me

# Simulate Data
df <- data.frame(year = sample(2005:2015, 20, replace = T),
                 crashes = sample(30:100, 20), 
                 type = sample(c('Vehicle', 'Pedestrian'), 20, replace = TRUE))
# Required packages
require(plotly); require(tidyverse)
# ggplot2 plot definition
crashplot <- df %>%
  group_by(year, type) %>%
  summarise(summed_crashes = sum(crashes)) %>%
  ggplot(., aes(x = year, y = summed_crashes, fill = type)) +
  geom_bar(stat = 'identity') + 
  labs(x = 'Crash Year',
       y = 'Total Crashes',
       title = 'Total Crashes by Year and Mode Type',
       fill = "Mode Type")
# Show plot with plotly
ggplotly(crashplot)

If it does not work, it would be good to see reproducible data example for this problem.

edited Oct 24 '20 at 10:36

answered Oct 23 '20 at 18:37

J. Ring

333
1
6

1

This would work except that my data frame is such that 1 row = 1 crash record. So the total crashes is a total of all the records in the data set, by mode and year. I have nearly 40,000 rows, each representing 1 crash. I think this is why it's showing as "1". The code I wrote seems to be "smart" enough to bin the crashes by year and mode in the bar chart, but the tooltip needs some other call to display the sum of the count. That's where I am having the issue. Please see the edited post with a reproducible data example. – myClone Oct 23 '20 at 22:05
1

@myClone - so you could add additional aggregation step before you plot with `dplyr`. Adding this to edit. – J. Ring Oct 24 '20 at 10:36
yes indeed. The aggregation step was the missing element required! Thank you and your edited response was marked as the correct solution. – myClone Oct 26 '20 at 15:59

How to Change the Tooltip in ggplotly to show SUM of a Column in a Bar Chart

Create ggplot

1 Answers1