[R] lmRobMM vs rlm

[daniel schubert]

I am needing some expertise with regard
to the S-Plus  command lmRobMM and its R counterpart 
rlm(formula,data,method="MM")

I have used lmRobMM(formula,data) in S-Plus on the Stackloss data and 
obtained for my residuals

6.217777 1.150717 6.427946 8.174019 -0.6713005 -1.248641 -0.4236203 
0.5763797 -1.057899 0.3593823

         11        12        13        14       15        16         17 
     18        19       20
  0.9629239 0.4732042 -2.506497 -1.346176 1.344408 0.1432279 -0.3729551 
0.09646618 0.5861859 1.934193

         21
  -8.629863


  (Intercept)        V1        V2          V3
    -37.65246 0.7976856 0.5773405 -0.06706018


but when I use rlm(formula,data,method="MM") in R my residuals are

           1            2            3            4            5            6
   2.82002948  -2.29288550   3.78605126   7.23168396  -1.60941126  -2.18886365
            7            8            9           10           11           12
  -1.09082616  -0.09082616  -1.43350203  -0.32664493   0.68958988   0.15612729
           13           14           15           16           17           18
  -3.10081497  -1.43820259   2.20031311   0.86156818  -0.29869392   0.49171093
           19           20           21
   1.02517352   1.61796106 -10.50956979

(Intercept)    VAR00001    VAR00002    VAR00003
-41.5230433   0.9388404   0.5794524  -0.1129150

I am very worried by this discrepancy what have I done wrong?

I read in R:Robust Fitting of Linear Models: MM estimation is M-Estimation 
with Tukey's biweight initialized by a specific S-estimator
and so I alter my S-Plus code to

ct<-lmRobMM.robust.control(weight=c("Optimal","Bisquare"))
fit<-lmRobMM(formula=V4~V1+V2+V3,data=data, robust.control=ct)

and obtain for my residuals

         1        2        3        4          5         6          7 
   8         9        10        11        12       13        14
  6.132346 1.061665 6.342018 8.212728 -0.7022119 -1.244742 -0.3631877 
0.6368123 -1.001597 0.2162886 0.8524152 0.3242646 -2.64235 -1.407392

        15        16         17         18        19       20        21
  1.338706 0.1266635 -0.4053969 0.08936822 0.6175189 1.894162 -8.820719

(Intercept)        V1        V2          V3
    -37.35298 0.8107863 0.5425301 -0.07068073


then I try

ct<-lmRobMM.robust.control(weight=c("Bisquare","Bisquare"))
fit<-lmRobMM(formula=V4~V1+V2+V3,data=data, robust.control=ct)

and obtain

   1        2        3       4          5         6          7         8 
      9        10        11        12        13        14
  5.954287 0.880977 6.206231 8.16344 -0.7473658 -1.291963 -0.3966992 
0.6033008 -1.020411 0.1894031 0.8491936 0.3204805 -2.663977 -1.402163

        15        16         17        18        19       20        21
  1.392297 0.1723669 -0.3985709 0.1145995 0.6433127 1.882605 -8.908035

(Intercept)        V1        V2          V3
    -37.56484 0.8178879 0.5445969 -0.07331006


no matter what I try I cannot reproduce my R-fit?


thank you for the time you spend with this query
cheers
Daniel