[R] Linear Model with Discrete Data

[Bert Gunter]

Lorenzo:

1. This is a statistics question, not an R question.

2. Your statistical background appears inadequate  -- it looks like
Poisson regression, which would fall under "generalized linear
models". But it depends on how "discrete" discrete is (on some level,
all measurements are discrete, discretized to the resolution of the
measurement process).

3. So I would advise seeking local statistical help. Getting
statistical advice remotely over the internet (even on a proper forum
for statistical advice, which this is not) is fraught with hazard and
the risk of bad science (not due to incompetence or maliciousness;
just due to the possibilities of misunderstanding and confusion) --
imho only, of course.

Of course, feel free to reject this and proceed at your own risk.

Cheers,
Bert



On Thu, Jun 13, 2013 at 1:49 PM, Lorenzo Isella
<lorenzo.isella at gmail.com> wrote:
> Dear All,
> I am struggling with a linear model and an allegedly trivial data set.
> The data set does not consist of categorical variables, but rather of
> numerical discrete variables (essentially, they count the number of times
> that something happened).
> Can I still use a standard linear regression, i.e. something like lm(y~x)?
> I attach a small snippet that illustrates the difficulties that I am
> experiencing (I do not understand why R complains about a list()).
> Any suggestion is appreciated.
> The data file can be downloaded from
>
> http://db.tt/hEKv1wH2
>
> Cheers
>
> Lorenzo
>
>
> #####################################
>
> data <- read.csv("testData.csv", header=TRUE)
>
>
> data <- subset(data,select= -c (X100, X182))
>
>
> y <- data$X358
>
> z <- subset(data, select=-c(X358))
>
> myLM <- lm(y~z)
>
>
> #####################
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.



-- 

Bert Gunter
Genentech Nonclinical Biostatistics

Internal Contact Info:
Phone: 467-7374
Website:
http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm