On 2017-02-06 20:39-0800 John Baumgardner wrote: > Dear Alan, > > Today I downloaded and installed version 5.12.0 on my Windows 10 machine > under Cygwin. It installed with no glitches whatever. However, when I > recompiled and ran my graphics program, some of the plots had no color. I > got the following runtime warning: > Plplot Fortran Warning: plscmap1la: inconsistent sizes for intensity, coord1, > coord2, coord3, alpha, and/or alt_hue_path > which tipped me off that there was likely a problem with how I was calling > plscmap1la, which loads the color map. In checking the PLPLOT documentation > for version 5.12.0, I noticed that there was a new argument,/npts/, that had > not been present in the earlier version I had been using. I then checked the > example program in examples/f95, x30f.f90, and noticed that the call there to > plscmap1la surprisingly lacked that new argument /npts/. When I added that > argument in my program, I found that my program would not compile properly > with /npts/ present. I then found the cause of the earlier warning message. > It was that the dimension of the final argument, alt_hue_path, was incorrect. > When I fixed that problem and left out the /npts/ argument, the program > compiled with no complaints, the earlier warning message went away, but the > plots still had no color. It is my suspicion that somehow in the 5.12.0 > there is something wrong with routine plscmap1la and that it has to do with > that new argument. > > Would you mind checking this out for me?
Hi John: I hope you don't mind I put your question and my response on the list because I think other Fortran users will be interested. Yes, the new Fortran binding (as we mentioned in the release notes) is now much more careful about the consistency of array sizes passed in a given call to one of the subroutines/functions in our Fortran API. Also, to be specific, plscmap1la is documented at <http://plplot.sourceforge.net/docbook-manual/plplot-html-5.12.0/plscmap1la.html>, but you should be aware that our Fortran binding (in fact all our supported languages other than C and C++) uses the redacted form of our API where redundant dimension arguments (for those languages since that same dimension information is carried in the arrays themselves) are dropped. In this case, the npts argument is one of those redacted arguments as can be seen from near the end of the above webpage where the redacted form of the plscmap1la is documented (with npts dropped). So it sounds to me like you are doing everything right (using the redacted API with consistent dimension information now for each of the array arguments to this routine). Yet you are still not getting any color. To pursue that issue further, could you please try building and running Fortran standard example 30 (which works well for me on Debian Jessie) on your Cygwin platform? To do that, use the -DBUILD_TEST=ON option for cmake, and build Fortran example 30 (and all its prerequisites) as follows: make x30f95 Also, build your favorite Cygwin PLplot device driver (I am going to assume that is svg). make svg Then run the example (in the build tree) using that device examples/f95/x30f -dev svg -o test.svg Then view that test.svg result with an svg viewer. (If you don't have such a specific svg viewer, your browser normally should be able to view svg images). If those results look good (i.e., have color), then you should try to see in what way your application differs from examples/f95/x30.f90 in the source tree. I hope these suggestions help you to find a quick solution to the no-color issue you have encountered with plplot-5.12.0. Alan __________________________ Alan W. Irwin Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the FreeEOS equation-of-state implementation for stellar interiors (freeeos.sf.net); the Time Ephemerides project (timeephem.sf.net); PLplot scientific plotting software package (plplot.sf.net); the libLASi project (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project (lbproject.sf.net). __________________________ Linux-powered Science __________________________ ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot _______________________________________________ Plplot-general mailing list Plplot-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-general
