Sigh. I never load packages in .Rprofile to avoid the irreproducibility trap. Might seem drastic to some, but I don't feel much pain because I almost always edit my code in a file rather than on the fly at the console, and re-run it frequently from a fresh R process to check my progress. -- Sent from my phone. Please excuse my brevity. On June 14, 2017 3:27:15 PM PDT, David Winsemius <dwinsemius at comcast.net> wrote:> >> On Jun 14, 2017, at 1:53 PM, Rolf Turner <r.turner at auckland.ac.nz> >wrote: >> >> On 15/06/17 05:29, David Winsemius wrote: >>>> On Jun 14, 2017, at 10:18 AM, David Winsemius ><dwinsemius at comcast.net> wrote: >>>> >>>> >>>>> On Jun 14, 2017, at 9:46 AM, Jeff Newmiller ><jdnewmil at dcn.davis.ca.us> wrote: >>>>> >>>>> I don't see a question. If your question is whether R supports >pattern fills, AFAIK it does not. If that is not your question, ask >one. >>>>> -- >>>>> Sent from my phone. Please excuse my brevity. >>>>> >>>>> On June 14, 2017 7:57:41 AM PDT, jean-philippe ><jeanphilippe.fontaine at gssi.infn.it> wrote: >>>>>> dear R users, >>>>>> >>>>>> I would like to fill a circle with yellow stripes instead of a >uniform >>>>>> yellow color. To draw the circle I used the following command >after >>>>>> having loaded the (very nice !) plotrix library : >>> I finally understood the question and it needs a hack to the >draw.circle function in plotrix since the angle and density arguments >don't get passed in: >>> First get code for draw.circle: >>> ------ >>> draw.circle # then copy to console and edit >>> draw.circle2 <- function (x, y, radius, nv = 100, border = NULL, >col = NA, lty = 1, >>> density=NA, angle=45, lwd = 1 ) >>> { >>> xylim <- par("usr") >>> plotdim <- par("pin") >>> ymult <- getYmult() >>> angle.inc <- 2 * pi/nv >>> angles <- seq(0, 2 * pi - angle.inc, by = angle.inc) >>> if (length(col) < length(radius)) >>> col <- rep(col, length.out = length(radius)) >>> for (circle in 1:length(radius)) { >>> xv <- cos(angles) * radius[circle] + x >>> yv <- sin(angles) * radius[circle] * ymult + y >>> polygon(xv, yv, border = border, col = col, lty = lty, >density=density, angle=angle, >>> lwd = lwd) >>> } >>> invisible(list(x = xv, y = yv)) >>> } >>> Now run your call to pdf with draw.circle2 instead of draw.circle. >> >> This is just idle curiosity, since I'm not really able to contribute >anything useful, but I can't resist asking: When I try to run the OP's >code I get an error: >> >>> Error in alpha("red", 0.4) : could not find function "alpha". >> >> Why does this (apparently) not happen to anyone else? Why does the >universe pick on *me*? What is the function "alpha()"? Where is it to >be found? > >I discovered some time ago that I no longer needed to load the ggplot2 >package. I wasn't entirely happy to make this discovery since I stilll >cling to the old lattice style. Eventually I figgured out that it was >because one of packages that I load in my .Rprofile-file had changed >its imports. The `alpha` function I see is from ggplot2. Resistance is >futile. I've now been partially assimilated. > > >> >> Searching on "alpha" is of course completely unproductive; there are >far too many (totally irrelevant) instances. > > >> >> cheers, >> >> Rolf >> >> -- >> Technical Editor ANZJS >> Department of Statistics >> University of Auckland >> Phone: +64-9-373-7599 ext. 88276 > >David Winsemius >Alameda, CA, USA
> On Jun 14, 2017, at 5:52 PM, Jeff Newmiller <jdnewmil at dcn.davis.ca.us> wrote: > > Sigh. I never load packages in .Rprofile to avoid the irreproducibility trap. Might seem drastic to some, but I don't feel much pain because I almost always edit my code in a file rather than on the fly at the console, and re-run it frequently from a fresh R process to check my progress.Yes, <sigh>. But I am a long-time user of the rms/Hmisc combo, as well as the survival package, so near the top of my .Rprofile is: require(lattice) require(sos) require(rms) Should I be ashamed of that? I suppose I should, and I _am_ ashamed of some of the other stuff in there ....<delete>, <delete> ... and I've been meaning to address my manifold deficiencies w.r.t. irreproducibility by moving to RStudio, but I keep putting it off. -- David.> -- > Sent from my phone. Please excuse my brevity. > > On June 14, 2017 3:27:15 PM PDT, David Winsemius <dwinsemius at comcast.net> wrote: >> >>> On Jun 14, 2017, at 1:53 PM, Rolf Turner <r.turner at auckland.ac.nz> >> wrote: >>> >>> On 15/06/17 05:29, David Winsemius wrote: >>>>> On Jun 14, 2017, at 10:18 AM, David Winsemius >> <dwinsemius at comcast.net> wrote: >>>>> >>>>> >>>>>> On Jun 14, 2017, at 9:46 AM, Jeff Newmiller >> <jdnewmil at dcn.davis.ca.us> wrote: >>>>>> >>>>>> I don't see a question. If your question is whether R supports >> pattern fills, AFAIK it does not. If that is not your question, ask >> one. >>>>>> -- >>>>>> Sent from my phone. Please excuse my brevity. >>>>>> >>>>>> On June 14, 2017 7:57:41 AM PDT, jean-philippe >> <jeanphilippe.fontaine at gssi.infn.it> wrote: >>>>>>> dear R users, >>>>>>> >>>>>>> I would like to fill a circle with yellow stripes instead of a >> uniform >>>>>>> yellow color. To draw the circle I used the following command >> after >>>>>>> having loaded the (very nice !) plotrix library : >>>> I finally understood the question and it needs a hack to the >> draw.circle function in plotrix since the angle and density arguments >> don't get passed in: >>>> First get code for draw.circle: >>>> ------ >>>> draw.circle # then copy to console and edit >>>> draw.circle2 <- function (x, y, radius, nv = 100, border = NULL, >> col = NA, lty = 1, >>>> density=NA, angle=45, lwd = 1 ) >>>> { >>>> xylim <- par("usr") >>>> plotdim <- par("pin") >>>> ymult <- getYmult() >>>> angle.inc <- 2 * pi/nv >>>> angles <- seq(0, 2 * pi - angle.inc, by = angle.inc) >>>> if (length(col) < length(radius)) >>>> col <- rep(col, length.out = length(radius)) >>>> for (circle in 1:length(radius)) { >>>> xv <- cos(angles) * radius[circle] + x >>>> yv <- sin(angles) * radius[circle] * ymult + y >>>> polygon(xv, yv, border = border, col = col, lty = lty, >> density=density, angle=angle, >>>> lwd = lwd) >>>> } >>>> invisible(list(x = xv, y = yv)) >>>> } >>>> Now run your call to pdf with draw.circle2 instead of draw.circle. >>> >>> This is just idle curiosity, since I'm not really able to contribute >> anything useful, but I can't resist asking: When I try to run the OP's >> code I get an error: >>> >>>> Error in alpha("red", 0.4) : could not find function "alpha". >>> >>> Why does this (apparently) not happen to anyone else? Why does the >> universe pick on *me*? What is the function "alpha()"? Where is it to >> be found? >> >> I discovered some time ago that I no longer needed to load the ggplot2 >> package. I wasn't entirely happy to make this discovery since I stilll >> cling to the old lattice style. Eventually I figgured out that it was >> because one of packages that I load in my .Rprofile-file had changed >> its imports. The `alpha` function I see is from ggplot2. Resistance is >> futile. I've now been partially assimilated. >> >> >>> >>> Searching on "alpha" is of course completely unproductive; there are >> far too many (totally irrelevant) instances. >> >> >>> >>> cheers, >>> >>> Rolf >>> >>> -- >>> Technical Editor ANZJS >>> Department of Statistics >>> University of Auckland >>> Phone: +64-9-373-7599 ext. 88276 >> >> David Winsemius >> Alameda, CA, USADavid Winsemius Alameda, CA, USA
On 15/06/17 13:51, David Winsemius wrote:> >> On Jun 14, 2017, at 5:52 PM, Jeff Newmiller <jdnewmil at dcn.davis.ca.us> wrote: >> >> Sigh. I never load packages in .Rprofile to avoid the irreproducibility trap. Might seem drastic to some, but I don't feel much pain because I almost always edit my code in a file rather than on the fly at the console, and re-run it frequently from a fresh R process to check my progress. > > Yes, <sigh>. But I am a long-time user of the rms/Hmisc combo, as well as the survival package, so near the top of my .Rprofile is: > > require(lattice) > require(sos) > require(rms) > > Should I be ashamed of that? > > I suppose I should, and I _am_ ashamed of some of the other stuff in > there ....<delete>, <delete> ... and I've been meaning to address my > manifold deficiencies w.r.t. irreproducibility by moving to RStudio, > but I keep putting it off.This is getting *way* off topic ... but why does using RStudio help with the irreproducibility problem? I thought that RStudio just made it easier to point-and-click. For those who like doing that sort of thing. (I tend to believe the dictum that a GUI makes it easy to do easy things and impossible to do hard things.) cheers, Rolf -- Technical Editor ANZJS Department of Statistics University of Auckland Phone: +64-9-373-7599 ext. 88276