R help - Apr 2007 - how to estimate dose from respond given drc package result

Dear all,

I can use the very nice drc package (multdrc()) to model and plot a  
dataframe containing dose and response values. I can also use  
predict.drc() to yield response values given a dose. I need to do the  
opposite, estimate a dose given the response. The general predict  
documentation seems to say that this is possible, but it does not  
appear that predict.drc has that capability. This makes sense because  
it is easy to compute y values from x values, given the fitted  
equation, but it is hard to do the opposite.

It seems that I need to do some kind of successive approximation. Is  
there a package that takes a multdrc generated model and estimates  
dose given response?

Thank you all for your time,

John

Try this:

library(drc)
# get dose corresponding to response (y = 2)
model1 <- multdrc(rootl ~ conc, data = ryegrass)
f <- function(x, y) y - predict(model1, data.frame(conc = x))[1]
uniroot(f, c(0, 10), y = 2)


On 4/17/07, John McNeil <jam at xardas.com> wrote:
> Dear all,
>
> I can use the very nice drc package (multdrc()) to model and plot a
> dataframe containing dose and response values. I can also use
> predict.drc() to yield response values given a dose. I need to do the
> opposite, estimate a dose given the response. The general predict
> documentation seems to say that this is possible, but it does not
> appear that predict.drc has that capability. This makes sense because
> it is easy to compute y values from x values, given the fitted
> equation, but it is hard to do the opposite.
>
> It seems that I need to do some kind of successive approximation. Is
> there a package that takes a multdrc generated model and estimates
> dose given response?
>
> Thank you all for your time,
>
> John
>
> ______________________________________________
> R-help at stat.math.ethz.ch 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.
>

Wow, thanks, that does it perfectly

John

On Apr 17, 2007, at 5:51 PM, Gabor Grothendieck wrote:
> Try this:
>
> library(drc)
> # get dose corresponding to response (y = 2)
> model1 <- multdrc(rootl ~ conc, data = ryegrass)
> f <- function(x, y) y - predict(model1, data.frame(conc = x))[1]
> uniroot(f, c(0, 10), y = 2)
>
>
> On 4/17/07, John McNeil <jam at xardas.com> wrote:
>> Dear all,
>>
>> I can use the very nice drc package (multdrc()) to model and plot a
>> dataframe containing dose and response values. I can also use
>> predict.drc() to yield response values given a dose. I need to do the
>> opposite, estimate a dose given the response. The general predict
>> documentation seems to say that this is possible, but it does not
>> appear that predict.drc has that capability. This makes sense because
>> it is easy to compute y values from x values, given the fitted
>> equation, but it is hard to do the opposite.
>>
>> It seems that I need to do some kind of successive approximation. Is
>> there a package that takes a multdrc generated model and estimates
>> dose given response?
>>
>> Thank you all for your time,
>>
>> John
>>
>> ______________________________________________
>> R-help at stat.math.ethz.ch 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.
>>