[R] Cumulative Density Plots (Hmisc/lattice)

[Deepayan Sarkar]

On 1/17/06, Mike Bock <mbock at environcorp.com> wrote:
> I have been using the ECDF function in the Hmisc package to produce
> cumulative distribution function plots. The problem is that for small
> datasets the steps "look bad" (not my characterization but from the
> client). Is there a way to get the same information but smoothed? I have
> tried the densityplot (lattice), which gives a smoothed line, but this
> does not give the cumulative density.

Have you tried qqmath, perhaps with distribution=qunif? Except for the
transposition, the idea behind the plots are similar.

Deepayan
--
http://www.stat.wisc.edu/~deepayan/