[R] Extract values from a data frame

[Ista Zahn]

Hi Laura,

On Thu, Aug 11, 2011 at 7:01 AM, Lali <laurafe at gmail.com> wrote:
> Hi everyone,
> I have a data frame that looks *sort of* like this:
>
> name <- letters[1:5]
> signal.1 <- c("12", "bad signal", "noise", "10", "X")
> length.signal.1 <- 5:9
> intensity.signal.1 <- 3:7
> signal.2 <- c("13", "noise", "19.2", "X", "V")
> length.signal.2 <- 2:6
> intensity.signal.2 <- 1:5
> signal.3 <- c("NA", "15.4", "error", "NA", "17")
> length.signal.3 <- c("NA", 2, 3, "NA", 4)
> intensity.signal.3 <- c("NA",4, 5, "NA", 5)
>
> #(there are actually up to 16 signals and 50 names, but I made this short
> for the example)
>
> df <- data.frame(cbind(name, signal.1, length.signal.1, intensity.signal.1,
> signal.2,
>                       length.signal.2, intensity.signal.2, signal.3,
> length.signal.3,
>                       intensity.signal.3))
>
>
>
> I need to "fish out" some values and have them in a new data frame.
>
> I am only interested in values in columns 2, 5 and 8 (actually seq(2, 50, 3)
> in my real df)
> I want the values that are not:
> "bad signal"
> "noise"
> "error"
> "NA"
> "V"
>
> This is the output I want (the name column is unimportant for my purposes,
> its just there as a reference for the example).
>
> (name)  S1       S2
> A        12        13
> B        15.4     (another value found in the other signals >3 not shown on
> example)
> C        19.2     (another value found in the other signals >3 not shown on
> example)
> D        10        X
> E        X         17
>
> I do know that there will always be 2 values exactly that do not match the
> exclusions named above, or none at all
>
> I have tried different approaches, grep, matching,%nin%... But as I am not
> an advanced used, I am very likely doing something wrong, because I either
> get a vector, or I get a matrix with TRUE FALSE, and usually I get the whole
> rows, and I don't want that :(
> I have also being searching the list for answers without avail.
> Any suggestions? Examples including syntax are appreciated (syntax is a
> major weak point for me).

Here is a solution using the reshape and plyr packages

library(reshape)
dfm <- melt(df[c(1, 2, 5, 8)], id = 1)
dfm.r <- dfm[!dfm$value %in% c("bad signal", "noise", "error", "NA", "V"), ]
dfm.r <- ddply(dfm.r, .(name), transform, index = paste("S",
1:length(name), sep = ""))
cast(dfm.r, name ~ index)

Best,
Ista
>
>
> Laura
>
>        [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org 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.
>



-- 
Ista Zahn
Graduate student
University of Rochester
Department of Clinical and Social Psychology
http://yourpsyche.org