Skip to content

From NUM to INT

7 messages · André Luis Neves, Jeff Newmiller, jim holtman +1 more

André Luis Neves
# 
Dear all:

I converted the columns (Baci, Meti, Fungii, Protozoai) into integers
(using excel) and then imported the data (.txt) into R. Interestingly, the
other three variables were loaded as INT, but the 'Baci' one continued as
Num.

I imported the data using the following command line:

X <- read.delim(file.choose(),
                 header = TRUE,
                 dec = ".")

Here is the structure of X:
'data.frame': 115 obs. of  5 variables:
 $ ID       : Factor w/ 61 levels "107ZRED","112BLKW",..: 8 12 15 18 26 27
29 31 32 36 ...
 $ Baci     : num  2.90e+12 5.55e+11 9.46e+11 8.13e+11 4.06e+11 ...
 $ Meti     : int  352645997 334146268 767208656 171567266 462747405
414905627 237010514 387480048 214671355 328813226 ...
 $ Fungii   : int  43645 19009 15998 2189 8972 8240 3133 17922 6156 13746
...
 $ Protozoai: int  3220523 1851891 3252462 1665675 34123768 23175015 203685
4261780 43110492 69802572 ...


I need Baci as an integer, and tried to convert it using as.integer
function, but was not successful.


Could anyone please help me to solve this problem.

Thanks,


?
Bert Gunter
# 
No.
"Note that current implementations of R use 32-bit integers for
integer vectors, so the range of representable integers is restricted
to about +/-2*10^9: doubles can hold much larger integers exactly. "

Cheers,
Bert


Bert Gunter

"The trouble with having an open mind is that people keep coming along
and sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )
On Tue, Apr 26, 2016 at 10:11 AM, Andr? Luis Neves <andrluis at ualberta.ca> wrote:
Jeff Newmiller
# 
That is the "impossible" case, since R integers are 32 bit signed (~2?10^9) even in 64 bit R. You can Google for "R arbitrary precision" and look for packages like Ryacas, bit64 or gmp. However, having such large integers stored as integers would not be necessary for most statistical analyses so you should confirm that your analysis cannot be performed using floating point numbers before resorting to that.
jim holtman
# 
Can you explain why you need them as 'integer',  A floating point
representation can hold a value upto ~4.5e15 as an "integer" keeping the
precision that you might need.


Jim Holtman
Data Munger Guru

What is the problem that you are trying to solve?
Tell me what you want to do, not how you want to do it.

On Tue, Apr 26, 2016 at 1:11 PM, Andr? Luis Neves <andrluis at ualberta.ca>
wrote:

  
  


  


      

    

    

      
      

        


  

        

      André Luis Neves
    

    

      
      #
    

  


  

      
Ok. I`m trying to run a Poisson glmm with an observation-level random
intercept. But I`m getting the following error for the 'Baci' variable:

'Error: (maxstephalfit) PIRLS step-halvings failed to reduce deviance in
pwrssUpdate'. I  guess this message is because the baci variable is not a
an integer, and cannot be transformed into an integer as R has a threshold
of
2x10^9 even in 64 bit R.

It runs fine for the fungii variable.

If you guys want to run the data (attached), the full command is below.

Thanks.

---------------------------------------------

##Import data:

qPCR <- read.delim(file.choose(),
                 header = TRUE,
                 dec = ".")

##Load package

library(lme4)

##Other steps:

qPCR$obs <- 1:nrow(qPCR)
qPCR$fID<-as.factor(qPCR$ID)
qPCR$fDiet<-as.factor(qPCR$Diet)

##Run the model:

M1 <- glmer (Baci ~ fDiet + Crossover + (1|fID:Crossover) + (1|obs),
             family = poisson, data=qPCR)



Andre

        
On Tue, Apr 26, 2016 at 2:36 PM, jim holtman <jholtman at gmail.com> wrote:

        

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Bert Gunter
    

    

      
      #
    

  


  

      
Inline.

-- Bert

        
On Tue, Apr 26, 2016 at 2:25 PM, Andr? Luis Neves <andrluis at ualberta.ca> wrote:

            

      
      
      
    

        
Why would you "guess" that? Given your magnitudes, I suspect it's a
scaling issue. Try centering and scaling your Baci variable before
fitting.

Cheers,
Bert





and cannot be transformed into an integer as R has a threshold

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

  
  


  


      

    

    

      
      

        


  

        

      André Luis Neves
    

    

      
      #
    

  


  

      
Hi Bert:

I thought in centering Baci variable, but I wouldn`t like to draw a
conclusion based on the scaled response variable.

On the other hand, if I had to center the explanatory variable, that would
be great, but I cannot, as it is a factor.

Thanks a lot for your time. I really appreciated your help.

Best regards,

Andre

        
On Tue, Apr 26, 2016 at 4:11 PM, Bert Gunter <bgunter.4567 at gmail.com> wrote: