R help - Aug 2017 - Plotting log transformed predicted values from lme

Hi,

I am performing meta-regression using linear mixed-effect model with the
lme() function  that has two fixed effect variables;one as a log
transformed variable (x)  and one as factor (y) variable, and two nested
random intercept terms.

I want to save the predicted values from that model and show the log curve
in a plot ; predicted~log(x)

mod<-lme(B~log(x)+as.factor(y), random=~1|cohort/Study,
weights=varFixed(~I(SE^2)), na.action=na.omit, data=subset(meta),
          control = lmeControl(sigma = 1, apVar = FALSE))
summary(mod)

newdat <- data.frame(x=seq(min(meta$x), max(meta$x),,118))  # I have 118
observations. #How do I add the factor variable to my newdat?
newdat$pred <- predict(mod, newdat,level = 0,type="response")

plot(B ~ x, data=meta)
lines(B ~ x, data=newdat)

Can you please assist me ?

Thank you!

Alina

	[[alternative HTML version deleted]]

Dear Alina

If I understand you correctly you cannot just have a single predicted 
curve but one for each level of your factor.

On 09/08/2017 16:24, Alina Vodonos Zilberg wrote:
> Hi,
>
> I am performing meta-regression using linear mixed-effect model with the
> lme() function  that has two fixed effect variables;one as a log
> transformed variable (x)  and one as factor (y) variable, and two nested
> random intercept terms.
>
> I want to save the predicted values from that model and show the log curve
> in a plot ; predicted~log(x)
>
> mod<-lme(B~log(x)+as.factor(y), random=~1|cohort/Study,
> weights=varFixed(~I(SE^2)), na.action=na.omit, data=subset(meta),
>           control = lmeControl(sigma = 1, apVar = FALSE))
> summary(mod)
>
> newdat <- data.frame(x=seq(min(meta$x), max(meta$x),,118))  # I have 118
> observations. #How do I add the factor variable to my newdat?
> newdat$pred <- predict(mod, newdat,level = 0,type="response")
>
> plot(B ~ x, data=meta)
> lines(B ~ x, data=newdat)
>
> Can you please assist me ?
>
> Thank you!
>
> Alina
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> 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.
>
> ---
> This email has been checked for viruses by AVG.
> http://www.avg.com
>
>
-- 
Michael
http://www.dewey.myzen.co.uk/home.html

Thank you Michael,
Curves for each level of the factor sounds  very interesting,
Do you have a suggestion how to plot them?

Thank you!
Alina

*Alina Vodonos Zilberg*




On Thu, Aug 10, 2017 at 7:39 AM, Michael Dewey <lists at
dewey.myzen.co.uk>
wrote:
> Dear Alina
>
> If I understand you correctly you cannot just have a single predicted
> curve but one for each level of your factor.
>
>
> On 09/08/2017 16:24, Alina Vodonos Zilberg wrote:
>
>> Hi,
>>
>> I am performing meta-regression using linear mixed-effect model with
the
>> lme() function  that has two fixed effect variables;one as a log
>> transformed variable (x)  and one as factor (y) variable, and two
nested
>> random intercept terms.
>>
>> I want to save the predicted values from that model and show the log
curve
>> in a plot ; predicted~log(x)
>>
>> mod<-lme(B~log(x)+as.factor(y), random=~1|cohort/Study,
>> weights=varFixed(~I(SE^2)), na.action=na.omit, data=subset(meta),
>>           control = lmeControl(sigma = 1, apVar = FALSE))
>> summary(mod)
>>
>> newdat <- data.frame(x=seq(min(meta$x), max(meta$x),,118))  # I have
118
>> observations. #How do I add the factor variable to my newdat?
>> newdat$pred <- predict(mod, newdat,level =
0,type="response")
>>
>> plot(B ~ x, data=meta)
>> lines(B ~ x, data=newdat)
>>
>> Can you please assist me ?
>>
>> Thank you!
>>
>> Alina
>>
>>         [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posti
>> ng-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>> ---
>> This email has been checked for viruses by AVG.
>> http://www.avg.com
>>
>>
>>
> --
> Michael
> http://www.dewey.myzen.co.uk/home.html
>
	[[alternative HTML version deleted]]

In the rockchalk package, I have a function called newdata that will help
with this. Plenty of examples. Probably my predictOmatic function will just
work. Motivation is in the vignette.

Paul Johnson
University of Kansas

On Aug 9, 2017 11:23 AM, "Alina Vodonos Zilberg" <alina.vodonos at
gmail.com>
wrote:
> Hi,
>
> I am performing meta-regression using linear mixed-effect model with the
> lme() function  that has two fixed effect variables;one as a log
> transformed variable (x)  and one as factor (y) variable, and two nested
> random intercept terms.
>
> I want to save the predicted values from that model and show the log curve
> in a plot ; predicted~log(x)
>
> mod<-lme(B~log(x)+as.factor(y), random=~1|cohort/Study,
> weights=varFixed(~I(SE^2)), na.action=na.omit, data=subset(meta),
>           control = lmeControl(sigma = 1, apVar = FALSE))
> summary(mod)
>
> newdat <- data.frame(x=seq(min(meta$x), max(meta$x),,118))  # I have 118
> observations. #How do I add the factor variable to my newdat?
> newdat$pred <- predict(mod, newdat,level = 0,type="response")
>
> plot(B ~ x, data=meta)
> lines(B ~ x, data=newdat)
>
> Can you please assist me ?
>
> Thank you!
>
> Alina
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> 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.
>
	[[alternative HTML version deleted]]