[R] Resistant Local Regression ???

[Maciej.Hoffman-Wecker@evotecoai.com]

Dear R-users & developers,


i'd like to know your opinion on the following suggestion.
In short:

Wouldn't it be possible to combine the properties of
local regression (loess) and resistant regression (lqs)
in order to cope with fitting of partially continuous data?

I think the lqs-fitting could detect and eliminate
the 2 outliers but perserve the discontinuity at t1
in case of the following data (hope its still courier).

 --------------------------------------------------
|                                                  |
|       *                                          |
|                                                  |
|                                                  |
|   *                                              |
| *  * *       * *                                 |
|        *    *                                    |
|          * *                                     |
|                                *                 |
|                                                  |
|                                                  |
|                                                  |
|                                                  |
|                                                  |
|                                      * *         |
|                 * *    *   * *     *     *       |
|                     * *  *      * *        * * * |
 -------|---------|-------------|------------------
        t0        t1            t2


Actually i'm interested in fitting a IR^2 -> IR
function. This is why i can't use lowess. (I guess
I shouldn't anyway - BTW, in a "lowess vs. loess"
mail Martin Maechler mentioned the robustness
properties of lowess ("huberizing"). Is lowess
more robust than loess with family=symmetric?.)

I will try to implement this in R-language,
in order to see, whether it works or not.

I hope to hear some opinion or recommendation.

Thanks in advance

maciej




Maciej Hoffman-Wecker
EVOTEC OAI
Screening Operations/Discovery Informatics
Schnackenburgallee 114
D-22525 Hamburg
Germany

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._