R help - May 2012 - Automating R for Hypothesis Testing

R Users-
  I have been trying to automate a manual code that I have developed for
calling in a .csv file, isolating certain rows and columns that correspond
to specified months:
something to the effect
i=name.csv
N=length(i$month)
iphos1=0
iphos2=0
isphos3=0
for i=1,N
 if month=1
    iphos1=iphos+1
    iphos1(iphos1)=i

an so on to call out the months into there own arrays (unless there is a way
I can wrap it into the next automation)

Next: I would like to run a simple linear regression combining each of the
months 1 by 1:
for instance I want to run a regression on a combined model from months 1
and 2 and a dummy model for 1 and 2, compare them using a Chi-sq
distribution, if Chi-sq is less than the Critical value, we accept and go on
to test another set of models with both 1 and 2. If it rejects, then we
proceed to months 2 and 3.  If we move on to the second set on months 1 and
2, and the critical value is accepted, I want to print an accept or reject
and move on to months 2 and 3, until finally comparing months 12-1 at the
end.
Is there a way to loop or automate this in R?

Thanks
Meredith


--
View this message in context:
http://r.789695.n4.nabble.com/Automating-R-for-Hypothesis-Testing-tp4618653.html
Sent from the R help mailing list archive at Nabble.com.

Hello,

I'm not at all sure if I understand your problem. Does this describe it?


test first model for months 1 and 2
if test statistic less than critical value{
	test second model for months 1 and 2
	print results of the first and second tests? just one of them?
}
move on to months 2 and 3
etc, until months 12 and 1


Please post example data using dput(dataset).
Just copy it's output and paste it in your post.

And example code, what you're already doing.
(Possibly simplified)

Rui Barradas


meredith wrote
> 
> R Users-
>   I have been trying to automate a manual code that I have developed for
> calling in a .csv file, isolating certain rows and columns that correspond
> to specified months:
> something to the effect
> i=name.csv
> N=length(i$month)
> iphos1=0
> iphos2=0
> isphos3=0
> for i=1,N
>  if month=1
>     iphos1=iphos+1
>     iphos1(iphos1)=i
> 
> an so on to call out the months into there own arrays (unless there is a
> way I can wrap it into the next automation)
> 
> Next: I would like to run a simple linear regression combining each of the
> months 1 by 1:
> for instance I want to run a regression on a combined model from months 1
> and 2 and a dummy model for 1 and 2, compare them using a Chi-sq
> distribution, if Chi-sq is less than the Critical value, we accept and go
> on to test another set of models with both 1 and 2. If it rejects, then we
> proceed to months 2 and 3.  If we move on to the second set on months 1
> and 2, and the critical value is accepted, I want to print an accept or
> reject and move on to months 2 and 3, until finally comparing months 12-1
> at the end.
> Is there a way to loop or automate this in R?
> 
> Thanks
> Meredith
> 

--
View this message in context:
http://r.789695.n4.nabble.com/Automating-R-for-Hypothesis-Testing-tp4618653p4619195.html
Sent from the R help mailing list archive at Nabble.com.