I would like to plot one piechart to represent 4 people, each has different values for 3 categories.
Name X Y Z
Bill 120 280 600
John 140 360 500
Johana 150 700 150
Tim 200 600 200
What I want to do is to have one pie-chart which is split into 4 equal areas (Bill,John, Johana, Tim), and per each area, it's split according to the values of X, Y, Z. I tried to plot values in R based on table, but unfortunately it does not work. Can anyone give me a guidance how this can be done in R ?
You need to reshape a bit your data to fit your needs. Use the reshape package to create a "melted" data frame:
df <- data.frame(
name=c('Bill','John','Johana','Tim'),
x = c(120,140,150,200),
y = c(280,360,700,600),
z = c(600,500,150,200)
)
library(reshape)
df2 <- sort_df(
melt(df, id='name')
, vars=c('name','variable')
)
pie(df2$value) # ONLY take the value column to create the chart.
I leave to you define the slices colors and labels.
The explanation:
The melt() function will "unpivot" your dataframe. The result will be something like this:
library(reshape)
melt(df, id='name')
name variable value
1 Bill x 120
2 John x 140
3 Johana x 150
4 Tim x 200
5 Bill y 280
6 John y 360
7 Johana y 700
8 Tim y 600
9 Bill z 600
10 John z 500
11 Johana z 150
12 Tim z 200
The sort_df() function will simply sort the entries by name and variable:
sort_df(melt(df, id='name'), vars=c('name','variable'))
name variable value
1 Bill x 120
5 Bill y 280
9 Bill z 600
3 Johana x 150
7 Johana y 700
11 Johana z 150
2 John x 140
6 John y 360
10 John z 500
4 Tim x 200
8 Tim y 600
12 Tim z 200
All that's left is plot the pie. As the sum of the values per name is 1000 for all the names, you'll not need to rescale the values. I leave to you how to rescale the values if the sum for each name is different.
References:
Collected from the Internet
Please contact [email protected] to delete if infringement.
Comments