R help - Jun 2010 - How to predict the mean and variance of the dependent variable after regression

Hi, folks,

As seen in the following codes:

x1=rlnorm(10)
x2=rlnorm(10,mean=2)
y=rlnorm(10,mean=10)### Fake dataset
linmod=lm(log(y)~log(x1)+log(x2))

After the regression, I would like to know the mean of y. Since log(y) is
normal and y is lognormal, I need to know the mean and variance of log(y)
first.  I tried mean (y) and mean(linmod), but either one is what I want.

Any tips?

Thanks in advance!

	[[alternative HTML version deleted]]

Hello,

If you just want the mean and variance of log(y) try:

mean(log(y))
var(log(y))

if there is missing data, you can add na.rm=TRUE to both of those.  If
you want the mean and variance of the predicted ys

mean(predict(linmod))
var(predict(linmod))

see

?mean
?var
?predict.lm #the specific method being used for predict() with model
objects of class lm

HTH,

Josh

On Mon, Jun 21, 2010 at 11:24 AM, Yi <liuyi.feier at gmail.com>
wrote:
> Hi, folks,
>
> As seen in the following codes:
>
> x1=rlnorm(10)
> x2=rlnorm(10,mean=2)
> y=rlnorm(10,mean=10)### Fake dataset
> linmod=lm(log(y)~log(x1)+log(x2))
>
> After the regression, I would like to know the mean of y. Since log(y) is
> normal and y is lognormal, I need to know the mean and variance of log(y)
> first. ?I tried mean (y) and mean(linmod), but either one is what I want.
>
> Any tips?
>
> Thanks in advance!
>
> ? ? ? ?[[alternative HTML version deleted]]
>
> ______________________________________________
> 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.
>


-- 
Joshua Wiley
Ph.D. Student
Health Psychology
University of California, Los Angeles

On Mon, Jun 21, 2010 at 11:45 AM, Yi <liuyi.feier at gmail.com>
wrote:
> Hi, Josh,
>
> Thank you very much! It is what I want!
> Because it is very obvious that the variance is not a constant in my linear
> model. So I am thinking about robust standand error. Any code works for
this
> purpose in R?
I do not have much experience in this area, but I do recall reading in
Venables and Ripley (Venables, W. N. & Ripley, B. D. (2002) Modern
Applied Statistics with S. Fourth Edition. Springer, New York. ISBN
0-387-95457-0) that they have a function for fitting robust linear
models in the package that goes with their book.

library(MASS) # load the package
?rlm # look at the help documentation
> BTW, It is very nice of you to tell me how to look up the function in R.
You are welcome.  You can also find a lot of information using the
RSiteSearch() function.  It can search the mailing list archives as
well as the documentation.
> Actually I don't understand?all the?information from summary(linmod).?I
> am?looking for books?for help. Please let me know,?if you?happen to know
> the?right?source for this.
There are several introductory books you could look at
http://www.r-project.org/doc/bib/R-books.html for a partial list.
Personally, I found Introductory Statistics with R by Peter Dalgaard
very helpful, but there are certainly others.

Best regards,

Josh
>
> Thank you again for your help :)
>
> Yi
>
> On Mon, Jun 21, 2010 at 11:34 AM, Joshua Wiley <jwiley.psych at
gmail.com>
> wrote:
>>
>> Hello,
>>
>> If you just want the mean and variance of log(y) try:
>>
>> mean(log(y))
>> var(log(y))
>>
>> if there is missing data, you can add na.rm=TRUE to both of those. ?If
>> you want the mean and variance of the predicted ys
>>
>> mean(predict(linmod))
>> var(predict(linmod))
>>
>> see
>>
>> ?mean
>> ?var
>> ?predict.lm #the specific method being used for predict() with model
>> objects of class lm
>>
>> HTH,
>>
>> Josh
>>
>> On Mon, Jun 21, 2010 at 11:24 AM, Yi <liuyi.feier at gmail.com>
wrote:
>> > Hi, folks,
>> >
>> > As seen in the following codes:
>> >
>> > x1=rlnorm(10)
>> > x2=rlnorm(10,mean=2)
>> > y=rlnorm(10,mean=10)### Fake dataset
>> > linmod=lm(log(y)~log(x1)+log(x2))
>> >
>> > After the regression, I would like to know the mean of y. Since
log(y)
>> > is
>> > normal and y is lognormal, I need to know the mean and variance of
>> > log(y)
>> > first. ?I tried mean (y) and mean(linmod), but either one is what
I
>> > want.
>> >
>> > Any tips?
>> >
>> > Thanks in advance!
>> >
>> > ? ? ? ?[[alternative HTML version deleted]]
>> >
>> > ______________________________________________
>> > 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.
>> >
>>
>>
>>
>> --
>> Joshua Wiley
>> Ph.D. Student
>> Health Psychology
>> University of California, Los Angeles
>
>


-- 
Joshua Wiley
Ph.D. Student
Health Psychology
University of California, Los Angeles