[R] Code translation from R to MATLAB

[Ana Cristiuc]

Dear Ted,



I have found on Internet your very good comments on R vs MATLAB languages.



I am a beginner in both languages, but unfortunately I have got a task to
translate a huge code from R to MATLAB.



Could you please recommend me some references or advices how to handle the
problems appearing from the differences between languages.



I was trying today to translate the below lines of code from R to MATLAB,
but very unsuccessfully.



Could you please help me by saying, at least, if I have any chances to
translate this R function to a function in  MATLAB or  it will be simpler
to avoid a function body for the initialization.



This is the first function from the attached script. It is the hardest part
of the code:





init <- function(N=3) {

    ## initialize model settings and load data

    ## Globals, side effects:

    ##  sets n.per, N, W, Y, dates, mats



    n.per <<- 12 # monthly data

    N <<- N      # three risk factors



    ## load yield data



    ## cat("loading GSW yield data...\n")

    ## if ('data' %in% dir()) {

    ##     load("data/gsw_data_monthly.RData")  # Y, dates, mats

    ## } else {

    ##     load("../data/gsw_data_monthly.RData")  # Y, dates, mats

    ## }

    ## ##mat.sel <- c(1:12)

    ## ##mat.sel <- c(3:12)  ## one through ten years

    ## mat.sel <- 2+c(1,2,3,4,5,7,10)

    ## Y <- Y[,mat.sel]/100/n.per

    ## mats <- mats[mat.sel]*n.per



    cat("loading Anh Le's yield data...\n")

    if ('data' %in% dir()) {

        load("data/le_data_monthly.RData")  # Y, dates

    } else {

        load("../data/le_data_monthly.RData")  # Y, dates

    }

    ## mats <- c(3,6,12,24,36,48,60,72,84,96,108,120)

    ##mats <- c(12,24,36,48,60,72,84,96,108,120)

    mats <- c(12,24,36,48,60,84,120)

    Y <- Y[,mats]/n.per



    if (dim(Y)[2]!=length(mats)) stop("wrong number of columns in yield
matrix");



    start.sample <- 19900101

    start.date <<- c(1990,1)

    ##start.sample <- 19850101

    ##start.date <<- c(1985,1)

    end.sample <- 20071231



    sel.sample <- dates>=start.sample & dates<=end.sample

    Y <- Y[sel.sample,]  ## change local variable

    dates <- dates[sel.sample]

    cat("sample from", as.character(min(dates)), "to",
as.character(max(dates)), "\n")



    ## first N principal components => pricing factors

    W <<- getW(Y, N)



    ## make local variables global   -- alt:  assign("Y", Y, env=.GlobalEnv)

    Y <<- Y; dates <<- dates; mats <<- mats

}



My understanding is that I can specify this initialization outside of a
function, but I am not sure.



I hope you will have some  free minutes to help me.



Many thanks in advance for your help!



Best regards,



Ana Cristiuc,