Help with vectorization

[Filoche]

Hi every one. I have a exponential function (3 fitting parameters) that I would like to use to produce data (6 series) without having to use a loop. Here

wl = seq(300,500,1)

k1 = c(1.2e-6, 4.9e-6, 9.6e-6, 2.7e-10, 6.7e-8, 7.44e-6)
k2 = c(726, 352, 128, 5232, 1538, 128)
k3 = c(-176, -224, -257, 88.7, -111, -256)

stations = c('R5d', 'R5a', 'R9', '108', '406', '409')


phiDIC = k1[1]*exp(k2[1]/(wl+k3[1])) #Here I would like to vectorize this function to create my 6 series instead of having to loop.


Regards,
Phil

[Michael Weylandt]

Perhaps ?outer -- well, not outer directly, but a multivariate outer
-- I keep this one around for personal use:

`mouter` <- function(..., FUN = "*"){
    dotArgs <- list(...)
    FUN <- match.fun(FUN)

    if(length(dotArgs) == 1L)
        return(unlist(dotArgs))
    if (length(dotArgs) == 2L)
        return(do.call(base::outer, c(dotArgs, FUN = FUN)))

        base::outer(dotArgs[[1]], do.call(Recall, c(dotArgs[-1], FUN = FUN)), FUN)
}

So then something like:

mouter(wl, k1, k2, k3, FUN = function(w, k1, k2, k3) k1 *exp(k2 / (w + k3)))

if I understand your request correctly.

Hope this helps,

Michael

On Thu, Apr 12, 2012 at 2:16 PM, Filoche <[hidden email]> wrote:

> Hi every one. I have a exponential function (3 fitting parameters) that I
> would like to use to produce data (6 series) without having to use a loop.
> Here
>
> wl = seq(300,500,1)
>
> k1 = c(1.2e-6, 4.9e-6, 9.6e-6, 2.7e-10, 6.7e-8, 7.44e-6)
> k2 = c(726, 352, 128, 5232, 1538, 128)
> k3 = c(-176, -224, -257, 88.7, -111, -256)
>
> stations = c('R5d', 'R5a', 'R9', '108', '406', '409')
>
>
> phiDIC = k1[1]*exp(k2[1]/(wl+k3[1])) #Here I would like to vectorize this
> function to create my 6 series instead of having to loop.
>
>
> Regards,
> Phil
>
> --
> View this message in context: http://r.789695.n4.nabble.com/Help-with-vectorization-tp4552638p4552638.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> [hidden email] 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.

______________________________________________
[hidden email] 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.

[Filoche]

Hi and thank you for your time.

I got this error when trying your function.

mouter(wl, k1, k2, k3, FUN = function(w, k1, k2, k3) k1 *exp(k2 / (w + k3)))
"Error in k3/(w + k3) : 'k3' is missing"

Regards,
Phil

[Michael Weylandt]

Strange, that isn't the error I get:

>  mouter(wl, k1, k2, k3, FUN = function(w, k1, k2, k3) k1 *exp(k2 / (w + k3)))
Error in FUN(X, Y, ...) : argument "k2" is missing, with no default

Still, it's a problem with my mouter() function which was only tested
on binary operators (and then only really to teach myself to use
Recall()). This should work better:

`mouter` <- function(..., FUN){
    dotArgs <-as.list(do.call("expand.grid", list(..., KEEP.OUT.ATTRS= F)))
    names(dotArgs) <- names(formals(FUN))
    ans <- do.call(FUN, dotArgs)
    dim(ans) <- sapply(list(...), length)
    ans
}

This only works when the function can take all the elements at a time
though; I'm sure some fooling around could combine them nicely...it
seems to test right on your data, but I haven't checked it more
generally.

Michael

On Thu, Apr 12, 2012 at 3:29 PM, Filoche <[hidden email]> wrote:

> Hi and thank you for your time.
>
> I got this error when trying your function.
>
> mouter(wl, k1, k2, k3, FUN = function(w, k1, k2, k3) k1 *exp(k2 / (w + k3)))
> "Error in k3/(w + k3) : 'k3' is missing"
>
> Regards,
> Phil
>
> --
> View this message in context: http://r.789695.n4.nabble.com/Help-with-vectorization-tp4552638p4552833.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> [hidden email] 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.

______________________________________________
[hidden email] 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.

[ilai-2]

Michael originally suggested ?outer. I think that was enough in this
case (no need for mv or sapply):

wlpk3 <- outer(k3,wl,'+')
ln.phiDIC <- log(k1)+k2/wlpk3
phiDIC<- t(exp(ln.phiDIC))
colnames(phiDIC)<- stations
str(phiDIC)

Cheers


On Thu, Apr 12, 2012 at 8:58 PM, R. Michael Weylandt
<[hidden email]> wrote:

> Strange, that isn't the error I get:
>
>>  mouter(wl, k1, k2, k3, FUN = function(w, k1, k2, k3) k1 *exp(k2 / (w + k3)))
> Error in FUN(X, Y, ...) : argument "k2" is missing, with no default
>
> Still, it's a problem with my mouter() function which was only tested
> on binary operators (and then only really to teach myself to use
> Recall()). This should work better:
>
> `mouter` <- function(..., FUN){
>    dotArgs <-as.list(do.call("expand.grid", list(..., KEEP.OUT.ATTRS= F)))
>    names(dotArgs) <- names(formals(FUN))
>    ans <- do.call(FUN, dotArgs)
>    dim(ans) <- sapply(list(...), length)
>    ans
> }
>
> This only works when the function can take all the elements at a time
> though; I'm sure some fooling around could combine them nicely...it
> seems to test right on your data, but I haven't checked it more
> generally.
>
> Michael
>
> On Thu, Apr 12, 2012 at 3:29 PM, Filoche <[hidden email]> wrote:
>> Hi and thank you for your time.
>>
>> I got this error when trying your function.
>>
>> mouter(wl, k1, k2, k3, FUN = function(w, k1, k2, k3) k1 *exp(k2 / (w + k3)))
>> "Error in k3/(w + k3) : 'k3' is missing"
>>
>> Regards,
>> Phil
>>
>> --
>> View this message in context: http://r.789695.n4.nabble.com/Help-with-vectorization-tp4552638p4552833.html
>> Sent from the R help mailing list archive at Nabble.com.
>>
>> ______________________________________________
>> [hidden email] 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.
>
> ______________________________________________
> [hidden email] 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.

______________________________________________
[hidden email] 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.

[Filoche]

Thank you all.

Problem solved.

Regards,
Phil