[R] Problem with Princurve

[David Winsemius]

On May 19, 2011, at 9:43 AM, guy33 wrote:

> Hey all,
>
> I can't seem to get the princurve package to produce correct  
> results, even
> in the simplest cases.  For example, if you just generate a 1 period
> noiseless sine wave, and ask for the principal curve and plot, the  
> returned
> curve is clearly wrong (doesn't follow the sine wave).  Here's my  
> code:
>
> library(princurve)
> x <- runif(1000,0,2*pi); x <- cbind(x/(2*pi), sin(x))
> fit1 <- principal.curve(x, plot = TRUE)
>
> Anyone have any suggestions?

The rule for contributed packages which you feel can be demonstrated  
to  give incorrect results is to contact the maintainer of the  
package. In this case I wonder if you even know what you are doing,  
however. Reading the help page for that function does not suggest to  
me that you should be expecting it to "go through the points"

"Fits a principal curve which describes a smooth curve that passes  
through the middle of the data x in an orthogonal sense."

The plotted curve does seem to adhere to that description with an  
emphasis on the last two words.

>  If you run this code, do you get the correct
> principal curve?

For the meaning assigned to "correct" by the authors, then yes.



David Winsemius, MD
West Hartford, CT