Skip to content

"Non-finite finite-difference value" error in eha's aftreg

6 messages · Milan Bouchet-Valat, Göran Broström, UyenThao Nguyen +2 more

Original
Milan Bouchet-Valat
# 
Hi list!


I'm getting an error message when trying to fit an accelerated failure
time parametric model using the aftreg() function from package eha:
This only happens when adding four specific covariates at the same time
in the model (see below). I understand that kind of problem can come
from a too high correlations between my covariates, but is there
anything I can do to avoid it? Does something need to be improved in
aftreg.fit?

My data set is constituted of 34,505 observations (years) of 2,717
individuals, which seems reasonable to me to fit a complex model like
that (covariates are all factors with less than 10 levels). I can send
it by private mail if somebody wants to help debugging this.

The details of the model and errors follow, but feel free to ask for
more testing. I'm using R 2.13.1 (x86_64-redhat-linux-gnu), eha 2.0-5
and survival 2.36-9.


Thanks for your help!
+                                    + t.since.school.q,
+            data=ms, dist="loglogistic", id=ident)
Error in optim(beta, Fmin, method = "BFGS", control = list(trace =
as.integer(printlevel)),  : 
  non-finite finite-difference value [2]
Calls: aftreg -> aftreg.fit -> aftp0 -> optim
4: optim(beta, Fmin, method = "BFGS", control = list(trace =
as.integer(printlevel)), 
       hessian = TRUE)
3: aftp0(printlevel, ns, nn, id, strata, Y, X, offset, dis, means)
2: aftreg.fit(X, Y, dist, strats, offset, init, shape, id, control, 
       center)
1: aftreg(Surv(start, end, event) ~ homo1 + sexego + dipref1 + 
       t.since.school.q, data = ms, dist = "loglogistic", id = ident)
14 days later
UyenThao Nguyen
# 
Hi All,

I've tried to install these "multtest" and "preprocessCore" packages in Mac, but kept getting error messages. I tried to load the packages using 2 ways:

1. Installed from BioConductor (sources)
2. And installed from BioConductor (binaries)

Both ways, I got these error messages:

For preprocessCore:
ld: warning: directory '/usr/local/lib' following -L not found
ld: library not found for -lgfortran
collect2: ld returned 1 exit status
make: *** [preprocessCore.so] Error 1
ERROR: compilation failed for package ?preprocessCore?
* removing ?/Library/Frameworks/R.framework/Versions/2.10/Resources/library/preprocessCore?
* restoring previous ?/Library/Frameworks/R.framework/Versions/2.10/Resources/library/preprocessCore?



For multtest:
ld: warning: directory '/usr/local/lib' following -L not found
** arch - x86_64


Looks like the error got to do with L drive, which I don't know how to fix it. Please help. Any suggestions will be greatly appreciated.

Thank you very much,
Nguyen
R. Michael Weylandt
# 
As the posting guide strongly suggests, I think the first step is to
update to R 2.14: there have been changes in package design and I'm
not sure back-compatibility will run that far. Doing so should let you
download fresh binaries straight from BioC.

I'm not sure why you would get compilation errors if you are just
installing a binary -- are you sure you're doing it right? What are
your commands?

I think follow-up should be directed to the dedicated BioConductor
list, but I assume they will tell you the same thing: 1) Update R and
try to install prepackaged libraries; 2) install local copies of the
binaries; 3) only compile from source if needed (you probably don't
have a fortran compiler which seems to be part of the problem)

Michael
On Wed, Nov 30, 2011 at 12:57 PM, UyenThao Nguyen <unguyen at tethysbio.com> wrote:
R. Michael Weylandt
# 
My apologies: the posting guide doesn't actually suggest updating to
the most recent stable version -- though perhaps it would be a good
addition to the "before posting" section -- but I still think that's
the necessary fix to your problem.

Michael

On Wed, Nov 30, 2011 at 1:08 PM, R. Michael Weylandt
<michael.weylandt at gmail.com> wrote:
David Winsemius
#
On Nov 30, 2011, at 12:57 PM, UyenThao Nguyen wrote:

            

      
      
      
    

        
You should read the directions about how to start with BioConductor.  
Generally, you need to install BiocLite or some such. I forget the  
exact specifics but they are all in the accompanying documentation.  
And if you need to install from sources, see comment below.

            

      
      
      
    

        
This suggests that in addition to not reading the Posting Guide (where  
you are asked not to append unrelated questions to unrelated threads   
AND you are asked to post Mac problems to the Mac list) that you also  
failed to read the Mac sections of the Installation and Administration  
Guide where you would have learned that you need to install Xcode.  
Please study that document more thouroughly if you plan to install  
from sources  ... although in this case that should not be necessary  
since the BioC repository has binaries.

            

      
      
      
    

        
There is no "L drive". Macs don't refer to their drives or partitions  
by letters. The dash indicates that "L" argument to a system command  
was not found.

            

      
      
      
    

        

      
      
    

        
David Winsemius, MD
West Hartford, CT