Skip to content

abline in log-log power law plot

3 messages · martin klein, Hadley Wickham, Peter Dalgaard

#
I have a power law plot in log-log scale using plot(x,y,log="xy") but I
can't get the regression line to plot correctly.
abline(lm(log(y)~log(x))) plots a straight line that looks promising but
the intercept with the y-axis seems to be way off.
summary(lm(log(y)~log(x))) gives the values but i cant really make any
sense out of it.
fitted(..) just confirms the values that seem to be incorrect.
any help is appreciated.
thanks!

martin
#
On Fri, Feb 22, 2008 at 2:38 PM, martin klein <mklein at cs.odu.edu> wrote:
You might try using ggplot2 instead:

install.packages("ggplot2")
library(ggplot2)

qplot(x, y, log="xy") + geom_smooth(method = lm)

Hadley
#
hadley wickham wrote:
Or just switch to base-10 logarithms:

abline(lm(log10(y)~log10(x)))