Error in seq.default(xmin, xmax, length.out = n) : 'from' must be a finite number.

[Enyu Tong]

Dear respected members of the INLA community,

I have been trying to use the BAPC/INLA package to project age-standardized rates. Below is the code I am using:

if(requireNamespace("INLA", quietly = TRUE)) {

    require(INLA)

    data(FemLCSweden)

    data(FemPYSweden)

    data(whostandard)

    lc_sweden <- APCList(FemLCSweden, FemPYSweden, gf=5)

    result  <- BAPC(lc_sweden, predict=list(npredict=10, retro=TRUE),

        verbose=FALSE, stdweight=whostandard[6:17,2])

    plotBAPC(result, scale=10^5, type="ageSpecProj", showdata=TRUE)

    ## Assume the data for 2000 to 2010 are not observed

    ## (projection results are the same as above)

    FemLCSweden[51:60,] <- NA

    lc_sweden <- APCList(FemLCSweden, FemPYSweden, gf=5)

    result  <- BAPC(lc_sweden, predict=list(npredict=10, retro=FALSE),

        verbose=FALSE, stdweight=whostandard[6:17,2])

    plotBAPC(result, scale=10^5, type="ageSpecProj", showdata=TRUE)

}

However, I keep encountering the error:

Error in seq.default(xmin, xmax, length.out = n) : 'from' must be a finite number.

I have tried multiple attempts to troubleshoot this issue, but it persists. Could you please guide how to resolve this error?

Best regards,

Enyu

[Finn Lindgren]

Hi,

First step is to identify what code is actually throwing the error, as it could be anywhere, including in a package unrelated to inla. The traceback() output after the error can help with that. The seq() function is an extremely commonly used R function…

Finn

On 16 Jun 2024, at 14:14, Enyu Tong <enyuto...@gmail.com> wrote:



--
You received this message because you are subscribed to the Google Groups "R-inla discussion group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to r-inla-discussion...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/r-inla-discussion-group/074d7697-0db1-425d-b8ab-eef14a6b4339n%40googlegroups.com.

[Helpdesk (Haavard Rue)]

I would guess there an Inf in the results (a marginal distribution) that cause
this, usually caused to large uncertainty that then sent through the exp()
function

Håvard Rue
he...@r-inla.org

[Helpdesk (Haavard Rue)]

your example runs fine w me. make sure you're using R-4.4 and a recent R-INLA
version