[R] Searching for specific values in a matrix

[Martin Maechler]

>>>>> "BertG" == Bert Gunter <gunter.berton at gene.com>
>>>>>     on Mon, 27 Jul 2009 15:30:08 -0700 writes:

    BertG> Nothing wrong with rolling your own, but see
    BertG> ?all.equal for R's built-in "almost.equal" version.

Yes, indeed!  Note that that needs a little extra care, as it
either returns TRUE or a character vector.
In sum, I'd strongly suggest you use

  %~% <- function(x,y) isTRUE(all.equal(x,y))

Martin Maechler, ETH Zurich and R Core Team

    BertG> Bert Gunter Genentech Nonclinical Biostatistics

    BertG> -----Original Message----- From:
    BertG> r-help-bounces at r-project.org
    BertG> [mailto:r-help-bounces at r-project.org] On Behalf Of
    BertG> Steve Lianoglou Sent: Monday, July 27, 2009 3:17 PM
    BertG> To: Mehdi Khan Cc: r-help at r-project.org Subject: Re:
    BertG> [R] Searching for specific values in a matrix

    BertG> Ahh ..

    BertG> On Jul 27, 2009, at 6:01 PM, Mehdi Khan wrote:

    >> Even when choosing a value from the first few rows, it
    >> doesn't work.  okay here it goes:
    >> 
    >> > rearranged[1:10, 1:5] x y band1 VSCAT.001 soiltype 1
    >> -124.3949 40.42468 NA NA CD 2 -124.3463 40.27358 NA NA CD
    >> 3 -124.3357 40.25226 NA NA CD 4 -124.3663 40.40241 NA NA
    >> CD 5 -124.3674 40.49810 NA NA CD 6 -124.3083 40.24744 NA
    >> 464 <NA> 7 -124.3017 40.31295 NA NA D 8 -124.3375
    >> 40.47557 NA 464 <NA> 9 -124.2511 40.11697 1 NA <NA> 10
    >> -124.2532 40.12640 1 NA <NA>
    >> 
    >> > query<- rearranged$y== 40.42468 > rearranged[query,]
    >> [1] x y band1 VSCAT.001 soiltype <0 rows> (or 0-length
    >> row.names)

    BertG> This isn't working because the numbers you see for y
    BertG> (40.42468) isn't precisely what that number is. As I
    BertG> mentioned before you should use an "almost.equals"
    BertG> type of search for this scenario. My "%~%" function
    BertG> isn't working in your session because that is a
    BertG> function I've defined myself. You can of course use
    BertG> it, you just have to define it in your
    BertG> workspace. Paste these lines into your workspace (or
    BertG> save them to a file and "source" that file into your
    BertG> workspace).

    BertG> ## === almost.equal functions ====

    BertG> almost.equal <- function(x, y,
    BertG> tolerance=.Machine$double.eps^0.5) { abs(x - y) <
    BertG> tolerance }

    BertG> "%~%" <- function(x, y) almost.equal(x, y)

    BertG> ## === end paste ==============

    BertG> Now you can use %~% once that's in. Let's use the
    BertG> almost.equal function now because I don't know if the
    BertG> default tolerance here is too strict (I suspect
    BertG> showing the value for rearranged$y[1] will show you
    BertG> more significant digits than you're seeing in the
    BertG> table(?))

    BertG> query <- almost.equal(rearranged$y, 40.42468,
    BertG> tolerance=0.0001) rearranged[query,]

    BertG> This will get you something.

    >> query<- rearranged$ VSCAT.001== 464 except it's a huge
    >> table (I guess I have to get rid of all rows with NA).

    BertG> Yes, I believe I mentioned earlier that you have to
    BertG> axe the NA matches manually:

    BertG> query <- rearranged$VSCAT.001 == 464 &
    BertG> !is.na(rearranged$VSCAT.001) rearranged[query,]

    BertG> Will get you what you want.

    >> I tried using the %~% but R doesn't recognize it.  So
    >> maybe it has to do with the rounding errors?

    BertG> Rounding errors won't happen with integer comparisons
    BertG> (and it looks like the VSCAT.001 columns is integers,
    BertG> no?).

    BertG> -steve

    BertG> -- Steve Lianoglou Graduate Student: Computational
    BertG> Systems Biology | Memorial Sloan-Kettering Cancer
    BertG> Center | Weill Medical College of Cornell University
    BertG> Contact Info: http://cbio.mskcc.org/~lianos/contact

    BertG> ______________________________________________
    BertG> R-help at r-project.org mailing list
    BertG> https://stat.ethz.ch/mailman/listinfo/r-help PLEASE
    BertG> do read the posting guide
    BertG> http://www.R-project.org/posting-guide.html and
    BertG> provide commented, minimal, self-contained,
    BertG> reproducible code.

    BertG> ______________________________________________
    BertG> R-help at r-project.org mailing list
    BertG> https://stat.ethz.ch/mailman/listinfo/r-help PLEASE
    BertG> do read the posting guide
    BertG> http://www.R-project.org/posting-guide.html and
    BertG> provide commented, minimal, self-contained,
    BertG> reproducible code.