CODEWITHSUNDEEP
rvisualizationrcharts
dfj328
dfj328

Reputation: 367

Drawing a custom chart in R - Displaying a time log of activites

I am new to R, so I would like to know if its possible to create a chart like the figure below?

enter image description here

I would like to create a custom chart displaying the duration of activities completed by various people at a given time.

Upvotes: 1

Views: 101

Answers (2)

bytesinflight
bytesinflight

Reputation: 1694

I'd like to propose another format for the data, where each activity has a start and an end time.

Load libraries. dplyr is only used for manipulating the dataset, and not strictly needed.

library(ggplot2)
library(dplyr)

First we read the data set

activities <- read.csv2(
text=
    "Person; Activity; Activity start; Activity end
    A; Activity 1; 10:30; 11:30
    A; Activity 2; 12:00; 13:00
    A; Activity 3; 14:00; 16:00
    B; Activity 1; 10:30; 11:30
    B; Activity 2; 11:30; 13:00
    B; Activity 3; 14:00; 14:30"
) %>%
    mutate(
        Activity.start = as.POSIXct(Activity.start, format="%H:%M"),
        Activity.end = as.POSIXct(Activity.end, format="%H:%M"),
        Person = as.factor(Person)
)

So now we have the correct classes for the columns and we can plot this with

ggplot(activities) +
    geom_rect(
        aes(
            xmin=Activity.start,
            xmax=Activity.end,
            fill=Activity,
            ymin=as.numeric(Person)-.5,
            ymax=as.numeric(Person)+.5)
        ) +
    scale_y_continuous(labels=levels(activities$Person), breaks=1:2) +
    geom_text(
        aes(x=(Activity.start + (Activity.end - Activity.start)/2), y=as.numeric(Person), label=Activity)
    ) +
    xlab(NULL) +
    ylab("Person")

which results in

Example plot

Upvotes: 1

Hack-R
Hack-R

Reputation: 23211

While this isn't normally the type of graph you'd create in R, surely you could pull it off with sufficient leg work.

I think this will get you jumpstarted:

require(lubridate)
require(ggplot2)

df              <- as.data.frame(matrix(nrow=14,ncol=3))
colnames(df)    <- c("Person", "Time", "Activity")
df$Person       <- c("A","A","A","A","A","A", "A", 
                     "B","B","B","B","B","B", "B")
df$Time         <- as.character(hours(c(10,11,12,13,14,15,16)))
df$Activity     <- c(NA, "Activity 1", NA, "Activity 2", NA, "Activity 3","Activity 3",
                     NA, "Activity 1", "Activity 2", NA, NA, "Activity 3", NA)  

ggplot(df, aes(Time, Person)) + geom_tile(aes(fill = Activity),
                                               colour = "white")  +   scale_fill_manual(values = rainbow (3))

enter image description here

You can take it to the next level by exploring the ggplot2 options and doing further customization to get it closer to the picture you drew. For instance you should be able to

  • change the dimensions of the bars
  • overlay the labels from the legend (geom_label is one way)
  • fine tune the tick marks

just by reading over and tweaking the options. I recently gave an answer on highly customized tick marks which you can find on SO with a quick search.

Upvotes: 1

Related Questions