Here's a thought, when I execute it on a data frame with only the subject
and resp variables I have the same problem. But not on the subset of data
we've been working with. The difference between the full vector and the
subset vector is that the full vector contains NA's. Only 35 out of 17,000
trials, but could this screw up the function?
unique(blah$resp)
[1] 2 1 NA 4
On Mon, Aug 8, 2011 at 11:46 AM, Joshua Wiley <jwiley.ps...@gmail.com>wrote:
> These are shots in the dark, but you could try: running it in a clean
> R session with only the necessary packages and data loaded; testing on
> subsets of your data; examing the classes of all your variables and
> make sure they are what is expexted; upgrading your version of R.
>
> I just retested this on:
>
> R version 2.12.1 (2010-12-16)
> Platform: x86_64-pc-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=English_United States.1252
> [2] LC_CTYPE=English_United States.1252
> [3] LC_MONETARY=English_United States.1252
> [4] LC_NUMERIC=C
> [5] LC_TIME=English_United States.1252
>
> attached base packages:
> [1] grid splines stats graphics grDevices utils datasets
> [8] methods base
>
> other attached packages:
> [1] car_2.0-9 survival_2.36-5 nnet_7.3-1 MASS_7.3-11
>
> again without issue, so I am thinking there must be something
> particular perhaps to your session going on, but it is difficult to
> say.
>
> On Mon, Aug 8, 2011 at 9:35 AM, Edward Patzelt <patze...@umn.edu> wrote:
> > Hmmmm, I pulled out a portion of the data set to create the code for
> > posting. When I execute this on the data frame I get the following for
> > subject 8 which is clearly incorrect. I get this for all subjects who
> > originally had "1 & 2".
> >
> > dat$respalt <- with(dat, ave(Slide1_RESP, factor(grid), FUN = foo))
> >> head(dat$respalt,20
> > + )
> > [1] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
> >
> > On Mon, Aug 8, 2011 at 11:14 AM, Joshua Wiley <jwiley.ps...@gmail.com>
> > wrote:
> >>
> >> Hmm, well I suppose it technically does "touch" in some sense still if
> >> there are 1 & 2s, but it should just return it as is, not changed.
> >> Thanks for the data, very easy! Here is what I get:
> >>
> >> dat <- structure(list(subject = c(8L, 8L, 8L, 8L, 8L, 8L, 8L, 8L, 8L,
> >> 8L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 6L, 6L,
> >> 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L), resp = c(2, 1, 1, 2, 1, 2, 1,
> >> 1, 1, 1, 4, 4, 2, 2, 4, 4, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 2, 2,
> >> 1, 2)), .Names = c("subject", "resp"), row.names = c(1L, 2L,
> >> 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 4161L, 4162L, 4163L, 4164L,
> >> 4165L, 4166L, 4167L, 4168L, 4169L, 4170L, 166L, 167L, 168L, 169L,
> >> 170L, 171L, 172L, 173L, 174L, 175L), class = "data.frame")
> >>
> >> require(car)
> >> foo <- function(x) {
> >> if (any(grepl(4, x))) {
> >> x <- recode(x, "2 = 1; 4 = 2")
> >> }
> >> return(x)
> >> }
> >>
> >> dat$respalt <- with(dat, ave(resp, factor(subject), FUN = foo))
> >>
> >> ## which at least for me gives:
> >> structure(list(subject = c(8L, 8L, 8L, 8L, 8L, 8L, 8L, 8L, 8L,
> >> 8L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 6L, 6L,
> >> 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L), resp = c(2, 1, 1, 2, 1, 2, 1,
> >> 1, 1, 1, 4, 4, 2, 2, 4, 4, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 2, 2,
> >> 1, 2), respalt = c(2, 1, 1, 2, 1, 2, 1, 1, 1, 1, 2, 2, 1, 1,
> >> 2, 2, 1, 1, 1, 1, 2, 2, 1, 1, 1, 1, 2, 2, 1, 2)), .Names = c("subject",
> >> "resp", "respalt"), row.names = c(1L, 2L, 3L, 4L, 5L, 6L, 7L,
> >> 8L, 9L, 10L, 4161L, 4162L, 4163L, 4164L, 4165L, 4166L, 4167L,
> >> 4168L, 4169L, 4170L, 166L, 167L, 168L, 169L, 170L, 171L, 172L,
> >> 173L, 174L, 175L), class = "data.frame")
> >>
> >> Does that work for you? I am running:
> >>
> >> R Under development (unstable) (2011-07-30 r56564)
> >> Platform: x86_64-pc-mingw32/x64 (64-bit)
> >> with car_2.0-10
> >>
> >> HTH,
> >>
> >> Josh
> >>
> >> On Mon, Aug 8, 2011 at 9:02 AM, Edward Patzelt <patze...@umn.edu>
> wrote:
> >> > Here's the code. I don't want it to even touch the vector if there
> are
> >> > already "1's & 2's"
> >> >
> >> > structure(list(subject = c(8L, 8L, 8L, 8L, 8L, 8L, 8L, 8L, 8L,
> >> > 8L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 42L, 6L, 6L,
> >> > 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L), resp = c(2, 1, 1, 2, 1, 2, 1,
> >> > 1, 1, 1, 4, 4, 2, 2, 4, 4, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 2, 2,
> >> > 1, 2)), .Names = c("subject", "resp"), row.names = c(1L, 2L,
> >> > 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 4161L, 4162L, 4163L, 4164L,
> >> > 4165L, 4166L, 4167L, 4168L, 4169L, 4170L, 166L, 167L, 168L, 169L,
> >> > 170L, 171L, 172L, 173L, 174L, 175L), class = "data.frame")
> >> >
> >> > R version 2.12.2 (2011-02-25)
> >> > Platform: x86_64-pc-mingw32/x64 (64-bit)
> >> >
> >> > locale:
> >> > [1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United
> >> > States.1252 LC_MONETARY=English_United States.1252
> >> > [4] LC_NUMERIC=C LC_TIME=English_United
> >> > States.1252
> >> >
> >> > attached base packages:
> >> > [1] splines stats graphics grDevices utils datasets
> methods
> >> > base
> >> >
> >> > other attached packages:
> >> > [1] car_2.0-10 nnet_7.3-1 MASS_7.3-11 Hmisc_3.8-3
> >> > survival_2.36-5 RODBC_1.3-2
> >> >
> >> > loaded via a namespace (and not attached):
> >> > [1] cluster_1.13.3 grid_2.12.2 lattice_0.19-17 tools_2.12.2
> >> >
> >> > On Mon, Aug 8, 2011 at 10:43 AM, Joshua Wiley <jwiley.ps...@gmail.com
> >
> >> > wrote:
> >> >>
> >> >> On Mon, Aug 8, 2011 at 8:23 AM, Edward Patzelt <patze...@umn.edu>
> >> >> wrote:
> >> >> > Thanks Josh for the code to post I have been trying to figure out
> how
> >> >> > to
> >> >> > do
> >> >> > that. Your code works except that it changes subjects that
> responded
> >> >> > with
> >> >> > "1 & 2" to all 1's. What does the "ave" argument mean in the
> >> >> > execution
> >> >> > of
> >> >>
> >> >> Not in the example data you provided on my system. If you can
> provide
> >> >> data (preferablly using dput() or uploading a txt file on a file
> >> >> hosting service) that reproduces this issue, I will be happy to look
> >> >> at it for you. You might also try reporting your sessionInfo() ---
> >> >> this may be related to the version of R or the packages you are
> using,
> >> >> but at present I have no information.
> >> >>
> >> >> > the function?
> >> >>
> >> >> 'ave' is not an argument in the function. ave() *is* a function. I
> >> >> call with() to have the ave() function evaluated in an environment
> >> >> created from dat (the data). See ?ave and ?with I could have
> >> >> equivalently (though more cumbersomely) written:
> >> >>
> >> >> ave(dat$Slide1_RESP, dat$Subject, FUN = foo))
> >> >>
> >> >> because ave will now be evaluated in the global environment, it will
> >> >> not have access to the variables stored in 'dat' unless explicitly
> >> >> told that they are in dat (as above).
> >> >>
> >> >> Cheers,
> >> >>
> >> >> Josh
> >> >>
> >> >> > library(car)
> >> >> > foo <- function(x) {
> >> >> > if (any(grepl(4, x))) {
> >> >> > x <- recode(x, "2 = 1; 4 = 2")
> >> >> > }
> >> >> > return(x)
> >> >> > }
> >> >> > ## do it
> >> >> > dat$test <- with(dat, ave(Slide1_RESP, Subject, FUN = foo))
> >> >> > On Fri, Aug 5, 2011 at 4:56 PM, Joshua Wiley <
> jwiley.ps...@gmail.com>
> >> >> > wrote:
> >> >> >>
> >> >> >> On Fri, Aug 5, 2011 at 2:34 PM, Edward Patzelt <patze...@umn.edu>
> >> >> >> wrote:
> >> >> >> >
> >> >> >> > The problem is that we were using a task where some subjects
> >> >> >> > responded
> >> >> >> > with "1 & 2" and some responded with "2 & 4". So there is
> overlap
> >> >> >> > for 2
> >> >> >> > because it means stimulus 1 for subject 1 and it means stimulus
> 2
> >> >> >> > for
> >> >> >> > subject 2.
> >> >> >> >
> >> >> >> > subject
> >> >> >> >
> >> >> >> > subject_1
> >> >> >> >
> >> >> >> > 1
> >> >> >> >
> >> >> >> > subject_1
> >> >> >> >
> >> >> >> > 1
> >> >> >> >
> >> >> >> > subject_1
> >> >> >> >
> >> >> >> > 1
> >> >> >> >
> >> >> >> > subject_1
> >> >> >> >
> >> >> >> > 2
> >> >> >> >
> >> >> >> > subject_1
> >> >> >> >
> >> >> >> > 2
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 4
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 2
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 2
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 4
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 2
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 2
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 2
> >> >> >> >
> >> >> >> > subject_2
> >> >> >> >
> >> >> >> > 4
> >> >> >>
> >> >> >> FYI providing data in the format above (this list is plain text)
> >> >> >> makes
> >> >> >> the job of those trying to help substantially harder. You can use
> >> >> >> dput(). For example, if I wanted to share the first 10 rows of
> the
> >> >> >> built in mtcars data set, I would just copy and paste the output
> >> >> >> from
> >> >> >> running:
> >> >> >>
> >> >> >> dput(mtcars[1:10, ])
> >> >> >>
> >> >> >> Anyway, here you go, this should be directly executable as long as
> >> >> >> you
> >> >> >> have installed the 'car' package.
> >> >> >>
> >> >> >> ## your data in a form easily copied and pasted into the console
> >> >> >> ## created using dput() (highly recommended for future posts)
> >> >> >> dat <- structure(list(subject = c("subject_1", "subject_1",
> >> >> >> "subject_1",
> >> >> >> "subject_1", "subject_1", "subject_2", "subject_2", "subject_2",
> >> >> >> "subject_2", "subject_2", "subject_2", "subject_2", "subject_2"
> >> >> >> ), val = c(1, 1, 1, 2, 2, 4, 2, 2, 4, 2, 2, 2, 4)), .Names =
> >> >> >> c("subject",
> >> >> >> "val"), row.names = c(NA, -13L), class = "data.frame")
> >> >> >>
> >> >> >> ## load the required package for recoding
> >> >> >> ## though it is overkill for only two levels
> >> >> >> require(car)
> >> >> >>
> >> >> >> ## define a function to do the recoding
> >> >> >> foo <- function(x) {
> >> >> >> if (any(grepl(4, x))) {
> >> >> >> x <- recode(x, "2 = 1; 4 = 2")
> >> >> >> }
> >> >> >> return(x)
> >> >> >> }
> >> >> >>
> >> >> >> ## do it
> >> >> >> dat$altval <- with(dat, ave(val, subject, FUN = foo))
> >> >> >>
> >> >> >> Cheers,
> >> >> >>
> >> >> >> Josh
> >> >> >>
> >> >> >>
> >> >> >> >
> >> >> >> > On Fri, Aug 5, 2011 at 4:25 PM, Joshua Wiley
> >> >> >> > <jwiley.ps...@gmail.com>
> >> >> >> > wrote:
> >> >> >> >>
> >> >> >> >> Hi Edward,
> >> >> >> >>
> >> >> >> >> You can try something like:
> >> >> >> >>
> >> >> >> >> u.ppl <- unique(init.dat1$grid)
> >> >> >> >> l.ppl <- ifelse(grepl(4, init.dat1$Slide1_RESP), 2,
> >> >> >> >> init.dat1$Slide1_RESP)
> >> >> >> >>
> >> >> >> >> Note that this is not exact as you have not provided a
> >> >> >> >> reproducible
> >> >> >> >> example. I am not exactly sure how you are putting 1 for 2
> and
> >> >> >> >> 2
> >> >> >> >> for
> >> >> >> >> 4, if the value is equal to 4, but presumably it is clearer
> with
> >> >> >> >> data.
> >> >> >> >> In any event, look at ?ifelse it is something like a
> vectorized
> >> >> >> >> if
> >> >> >> >> statement and is, I believe, preferable to your use of a for
> >> >> >> >> loop.
> >> >> >> >> I
> >> >> >> >> can probably give you a runnable solution if you can give the
> >> >> >> >> first
> >> >> >> >> few rows of the relevant data.
> >> >> >> >>
> >> >> >> >> Cheers,
> >> >> >> >>
> >> >> >> >> Josh
> >> >> >> >>
> >> >> >> >> On Fri, Aug 5, 2011 at 2:15 PM, Edward Patzelt <
> patze...@umn.edu>
> >> >> >> >> wrote:
> >> >> >> >> > I want to write code that says "If you find an element equal
> to
> >> >> >> >> > 4
> >> >> >> >> > in
> >> >> >> >> > this
> >> >> >> >> > vector for each person in the data set tested separately,
> then
> >> >> >> >> > put
> >> >> >> >> > in
> >> >> >> >> > 1 for
> >> >> >> >> > 2 and 2 for 4, else leave the variable as is"
> >> >> >> >> >
> >> >> >> >> > u.ppl <- (unique(init.dat1$grid))
> >> >> >> >> > l.ppl <- length(u.ppl)
> >> >> >> >> > for (i in 1:l.ppl)
> >> >> >> >> > {
> >> >> >> >> > if (grep("4",init.dat1$Slide1_RESP)) {2 == 1, 4 ==
> 2};
> >> >> >> >> > else
> >> >> >> >> > init.dat1$Slide1_RESP
> >> >> >> >> >
> >> >> >> >> > }
> >> >> >> >> >
> >> >> >> >> > --
> >> >> >> >> > Edward H. Patzelt
> >> >> >> >> > Research Assistant TRiCAM Lab
> >> >> >> >> > University of Minnesota Psychology/Psychiatry
> >> >> >> >> > VA Medical Center
> >> >> >> >> > Office: S355 Elliot Hall - Twin Cities Campus
> >> >> >> >> > Phone: 612-626-0072 Email: patze...@umn.edu
> >> >> >> >> >
> >> >> >> >> > Please consider the environment before printing this email
> >> >> >> >> > www.psych.umn.edu/research/tricam
> >> >> >> >> >
> >> >> >> >> > [[alternative HTML version deleted]]
> >> >> >> >> >
> >> >> >> >> >
> >> >> >> >> > ______________________________________________
> >> >> >> >> > R-help@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.
> >> >> >> >> >
> >> >> >> >> >
> >> >> >> >>
> >> >> >> >>
> >> >> >> >>
> >> >> >> >> --
> >> >> >> >> Joshua Wiley
> >> >> >> >> Ph.D. Student, Health Psychology
> >> >> >> >> Programmer Analyst II, ATS Statistical Consulting Group
> >> >> >> >> University of California, Los Angeles
> >> >> >> >> https://joshuawiley.com/
> >> >> >> >
> >> >> >> >
> >> >> >> >
> >> >> >> > --
> >> >> >> > Edward H. Patzelt
> >> >> >> > Research Assistant TRiCAM Lab
> >> >> >> > University of Minnesota Psychology/Psychiatry
> >> >> >> > VA Medical Center
> >> >> >> > Office: S355 Elliot Hall - Twin Cities Campus
> >> >> >> > Phone: 612-626-0072 Email: patze...@umn.edu
> >> >> >> >
> >> >> >> > Please consider the environment before printing this email
> >> >> >> > www.psych.umn.edu/research/tricam
> >> >> >>
> >> >> >>
> >> >> >>
> >> >> >> --
> >> >> >> Joshua Wiley
> >> >> >> Ph.D. Student, Health Psychology
> >> >> >> Programmer Analyst II, ATS Statistical Consulting Group
> >> >> >> University of California, Los Angeles
> >> >> >> https://joshuawiley.com/
> >> >> >
> >> >> >
> >> >> >
> >> >> > --
> >> >> > Edward H. Patzelt
> >> >> > Research Assistant TRiCAM Lab
> >> >> > University of Minnesota Psychology/Psychiatry
> >> >> > VA Medical Center
> >> >> > Office: S355 Elliot Hall - Twin Cities Campus
> >> >> > Phone: 612-626-0072 Email: patze...@umn.edu
> >> >> >
> >> >> > Please consider the environment before printing this email
> >> >> > www.psych.umn.edu/research/tricam
> >> >> >
> >> >>
> >> >>
> >> >>
> >> >> --
> >> >> Joshua Wiley
> >> >> Ph.D. Student, Health Psychology
> >> >> Programmer Analyst II, ATS Statistical Consulting Group
> >> >> University of California, Los Angeles
> >> >> https://joshuawiley.com/
> >> >
> >> >
> >> >
> >> > --
> >> > Edward H. Patzelt
> >> > Research Assistant TRiCAM Lab
> >> > University of Minnesota Psychology/Psychiatry
> >> > VA Medical Center
> >> > Office: S355 Elliot Hall - Twin Cities Campus
> >> > Phone: 612-626-0072 Email: patze...@umn.edu
> >> >
> >> > Please consider the environment before printing this email
> >> > www.psych.umn.edu/research/tricam
> >> >
> >>
> >>
> >>
> >> --
> >> Joshua Wiley
> >> Ph.D. Student, Health Psychology
> >> Programmer Analyst II, ATS Statistical Consulting Group
> >> University of California, Los Angeles
> >> https://joshuawiley.com/
> >
> >
> >
> > --
> > Edward H. Patzelt
> > Research Assistant TRiCAM Lab
> > University of Minnesota Psychology/Psychiatry
> > VA Medical Center
> > Office: S355 Elliot Hall - Twin Cities Campus
> > Phone: 612-626-0072 Email: patze...@umn.edu
> >
> > Please consider the environment before printing this email
> > www.psych.umn.edu/research/tricam
> >
>
>
>
> --
> Joshua Wiley
> Ph.D. Student, Health Psychology
> Programmer Analyst II, ATS Statistical Consulting Group
> University of California, Los Angeles
> https://joshuawiley.com/
>
--
Edward H. Patzelt
Research Assistant TRiCAM Lab
University of Minnesota Psychology/Psychiatry
VA Medical Center
Office: S355 Elliot Hall - Twin Cities Campus
Phone: 612-626-0072 Email: patze...@umn.edu
Please consider the environment before printing this email
www.psych.umn.edu/research/tricam
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
