R help - Oct 2010 - Strange glm(, quasipoisson) error

Dear list,

I have recently encountered an odd error when running glm(dep~indep,  
quasipoisson): while, with a subset of my data, I could get a  
perfectly reasonable model, once I include all of my data (17K+  
observations, 29 variables), I get the following error:

Error in if (any(y < 0)) stop("negative values not allowed for the  
quasiPoisson family") :
   missing value where TRUE/FALSE needed
In addition: Warning message:
In Ops.factor(y, 0) : < not meaningful for factors

I say this is odd because I triple checked my response variable, and  
not a single observation is negative. Is R encountering an error due  
to the size of the data, and somehow returning a 'random' error?  
(would be stranger still!)

Thanks a million, hope this makes sense.

Cheers,


Wil

PS.:  I tried the following solution
http://markmail.org/message/353xeiqtqj6jwxsx
, which basically rewrites the quasipoisson to 'obviate' negative  
values (when using that code, I assigned it to quasipoisson2, just in  
case). Either way, I got a different error, which suggests that  
something else is going on...

Error in if (!(validmu(mu) && valideta(eta))) stop("cannot find
valid
starting values: please specify some",  :
   missing value where TRUE/FALSE needed
In addition: Warning message:
In mean.default(y) : argument is not numeric or logical: returning NA

Don't know if it helps...


On Oct 19, 2010, at 00:55 , Wil M Contreras Arbaje wrote:
> Dear list,
>
> I have recently encountered an odd error when running glm(dep~indep,  
> quasipoisson): while, with a subset of my data, I could get a  
> perfectly reasonable model, once I include all of my data (17K+  
> observations, 29 variables), I get the following error:
>
> Error in if (any(y < 0)) stop("negative values not allowed for the
> quasiPoisson family") :
>  missing value where TRUE/FALSE needed
> In addition: Warning message:
> In Ops.factor(y, 0) : < not meaningful for factors
>
> I say this is odd because I triple checked my response variable, and  
> not a single observation is negative. Is R encountering an error due  
> to the size of the data, and somehow returning a 'random' error?  
> (would be stranger still!)
>
> Thanks a million, hope this makes sense.
>
> Cheers,
>
>
> Wil

Hi everyone,

Please ignore my previous message: it turns out Excel, when saving  
as .csv, was somehow storing numerical values as text, and that was  
causing the error. I opened the .csv, changed all observations to -->  
Number type, and it worked w/o a hitch afterwards.

Thanks,


Wil


On Oct 19, 2010, at 00:55 , Wil M Contreras Arbaje wrote:
> Dear list,
>
> I have recently encountered an odd error when running glm(dep~indep,  
> quasipoisson): while, with a subset of my data, I could get a  
> perfectly reasonable model, once I include all of my data (17K+  
> observations, 29 variables), I get the following error:
>
> Error in if (any(y < 0)) stop("negative values not allowed for the
> quasiPoisson family") :
>  missing value where TRUE/FALSE needed
> In addition: Warning message:
> In Ops.factor(y, 0) : < not meaningful for factors
>
> I say this is odd because I triple checked my response variable, and  
> not a single observation is negative. Is R encountering an error due  
> to the size of the data, and somehow returning a 'random' error?  
> (would be stranger still!)
>
> Thanks a million, hope this makes sense.
>
> Cheers,
>
>
> Wil