Reputation: 15
I am working with California Housing Dataset. The dataset has 20640 observations and 10 attributes. I am using R to make biplot but the figure I obtained is not very readable. The output is as followenter image description here
I am using a simple code to make this output.
biplot(housingpr,scale = 0)
Is there anyway to make this biplot look readble.
Upvotes: 0
Views: 679
Reputation: 11056
There is not much you can do if you want to plot 20,640 observations except to make the points smaller. Here is an example with the iris
data:
data(iris)
iris.pca <- prcomp(iris[, -5], scale.=TRUE)
biplot(iris.pca, xlabs=rep("*", nrow(iris)), cex=.75)
The xlabs=
argument sets the text for each point with the default value being the row number. This replaces the default with an asterisk for each value. If there are still too many points you can replace the asterisk with a period. The cex=
argument controls the size of the labels with the default value of 1 being full-size.
Upvotes: 2