R help - Jul 2013 - coxph doesn't converge when including factor variables

Hello,

 

I have a dataset of 371 observations.

When I run coxph with numeric variables it works fine.

However, when I try to add factor variables it returns "Ran out of
iterations and the model did not converge"

There is something very strange with the factors - some of them should
actually be numeric and all have a category of (other) with many cases
falling into this category. However, when looking at the data there is no
(other) category, nor are there empty/NA values.

I have tried casting the numeric "factors" into numeric using
as.numeric(levels(dataset$ABS_BLST))[as.integer(dataset$ABS_BLST)] without
success.

Of note, when I restructure all factors to binary variables with dummy and
use glmnet-lasso the model converges.

 

Here are examples of the code and output (including summary description of
the variables):
> maxSTree.cox <- coxph (Surv(time,status)~Chemo_Simple, data=dataset)
 

Warning message:

In fitter(X, Y, strats, offset, init, control, weights = weights,  :

  Ran out of iterations and did not converge

 
> summary (dataset$Chemo_Simple)
                        Anthra-HDAC       Anthra-Plus       ArsenicAtra
ATRA           ATRA-GO 

                0               163                 2                12
0                 2 

         ATRA_IDA Demeth-HistoneDAC          Flu-HDAC     Flu-HDAC-plus
HDAC-Clof         HDAC-only 

                0                34                37                 4
24                 1 

        HDAC-Plus        LowArac+/-       LowDAC-Clof         MYLO+IL11
No Rx in MDACC            Phase1 

                4                 8                30                 5
1                 5 

              SCT    StdARAC-Anthra      StdAraC-Plus          Targeted
VNP40101M 

                0                 0                 0                13
23 

 

Error 2:
> maxSTree.cox <- coxph (Surv(time,status)~ABS_BLST, data=dataset)
 

Error in fitter(X, Y, strats, offset, init, control, weights = weights,  : 

  NA/NaN/Inf in foreign function call (arg 6)

-----------------> for some reason ABS_BLST is a factor and not numeric.
Tried to cast it with as.numeric (didn't work)

 
> summary (dataset$ABS_BLST)
      0     195     153     160      17     192      25      26    2600
300      44      52      96     100 

     62       3       2       2       2       2       2       2       2
2       2       2       2       1 

   1020    1037     104  104076   10422   10441  106220   10887  111340
11154    1122   11250    1134   11388 

      1       1       1       1       1       1       1       1       1
1       1       1       1       1 

   1140   11427    1148   11577   11726    1176   11834    1188    1200
12012    1218   12222    1224    1232 

      1       1       1       1       1       1       1       1       1
1       1       1       1       1 

    128    1287  128860     129     130     132    1320    1326   13440
1347     136  136590   13728   13915 

      1       1       1       1       1       1       1       1       1
1       1       1       1       1 

     14   14112    1426   14406   14430   14712     148    1496     150
15184     152   15300  153538     156 

      1       1       1       1       1       1       1       1       1
1       1       1       1       1 

   1587   15884    1610    1617   16200   16300   16405    1680  168912
170520     171   17202   17340    1746 

      1       1       1       1       1       1       1       1       1
1       1       1       1       1 

   1750    1758    1767   17748    1800  180120    1813   18411    1860
18802    1900    1930    1950   19758 

      1       1       1       1       1       1       1       1       1
1       1       1       1       1 

   1980 (Other) 

      1     195
----------------------------> for some reason it ABS_BLST is a factor
listing (other) 195, but there are only numeric values (when I check the
dataset).   as.numeric (dataset$ABS_BLST did not do anything).

 

HELP !!!!


	[[alternative HTML version deleted]]