From: Michael H. <hag...@jp...> - 2003-02-03 16:39:51
|
Andrew, Andrew Straw <and...@ad...> wrote: > First of all, a big thanks for contributing Gnuplot.py! I use it > all the time, as I'm sure countless others do. Thanks for the kind words. > Now a question -- is there a way to pipe data through a > Gnuplot.PlotItems.TempArrayFile by default? I ask because on my > machine (Mac OS X 10.2.3 with gnuplot 3.8h0), the 'set mouse' on the > x11 terminal only works when the data is available in such a file. > > I vaguely seem to recall another version of Gnuplot (I'm using 1.6) did > this. (I remember lots of /tmp/@1234132.1 files.) I did find in the > gp_macosx.py file a "prefer_inline_data" variable, which I have set to > 1, to no avail. That's funny; I thought that in version 1.6 temporary files were still the default for Mac OS X. In any case, in the current CVS version, FIFOs are used by default to send data to gnuplot on platforms that support them, including Mac OS X. In the CVS version, if you set the following options in gp_macos.py, then you should get temporary files by default: prefer_inline_data = 0 prefer_fifo_data = 1 You can obtain the CVS version from sourceforge; instructions are available one or two clicks from the Gnuplot.py project page, http://sourceforge.net/projects/gnuplot-py/ Also from the sourceforge project page you can find instructions how to join our mailing list, which is a good place to submit such questions and discussion. (I have forwarded a copy of this email to the mailing list.) Michael -- Michael Haggerty hag...@jp... |
From: Andrew S. <and...@ad...> - 2003-02-04 14:08:05
|
It sounds good... However, I just did a CVS checkout, build, and install and got the following: >>> import Gnuplot Traceback (most recent call last): File "<stdin>", line 1, in ? File "/Library/Frameworks/Python.framework/Versions/2.2/lib/python2.2/site- packages/Gnuplot/__init__.py", line 177, in ? from PlotItems import PlotItem, Func, File, Data, GridData File "/Library/Frameworks/Python.framework/Versions/2.2/lib/python2.2/site- packages/Gnuplot/PlotItems.py", line 374, in ? if gp.GnuplotOpts.support_fifo: AttributeError: class GnuplotOpts has no attribute 'support_fifo' (I'll dig farther, but this is what I got to tonight.) Cheers! Andrew On Tuesday, February 4, 2003, at 03:09 AM, Michael Haggerty wrote: > Andrew, > > Andrew Straw <and...@ad...> wrote: >> First of all, a big thanks for contributing Gnuplot.py! I use it >> all the time, as I'm sure countless others do. > > Thanks for the kind words. > >> Now a question -- is there a way to pipe data through a >> Gnuplot.PlotItems.TempArrayFile by default? I ask because on my >> machine (Mac OS X 10.2.3 with gnuplot 3.8h0), the 'set mouse' on the >> x11 terminal only works when the data is available in such a file. >> >> I vaguely seem to recall another version of Gnuplot (I'm using 1.6) >> did >> this. (I remember lots of /tmp/@1234132.1 files.) I did find in the >> gp_macosx.py file a "prefer_inline_data" variable, which I have set to >> 1, to no avail. > > That's funny; I thought that in version 1.6 temporary files were still > the default for Mac OS X. > > In any case, in the current CVS version, FIFOs are used by default to > send data to gnuplot on platforms that support them, including Mac OS > X. In the CVS version, if you set the following options in > gp_macos.py, then you should get temporary files by default: > > prefer_inline_data = 0 > prefer_fifo_data = 1 > > You can obtain the CVS version from sourceforge; instructions are > available one or two clicks from the Gnuplot.py project page, > > http://sourceforge.net/projects/gnuplot-py/ > > Also from the sourceforge project page you can find instructions how > to join our mailing list, which is a good place to submit such > questions and discussion. (I have forwarded a copy of this email to > the mailing list.) > > Michael > > -- > Michael Haggerty > hag...@jp... > |
From: Andrew S. <and...@ad...> - 2003-02-05 04:32:17
|
Sorry, my error report was premature -- deleting my old Gnuplot module before re-installing fixes the problem. And I now have working mouse zoom! Unfortunately, I cannot get this to work with fifos -- it seems the "file" is unlinked immediately after gnuplot gets it the first time, and gnuplot wants it again when trying to zoom. Here's the output when I put debug=1: gnuplot> set mouse gnuplot> set term x11 gnuplot> set yrange [-5:20] gnuplot> set xlabel "Time (seconds)" gnuplot> set ylabel "Response (mV)" gnuplot> plot '/tmp/@20093.0' title "1115: 2000-03-06 17:23:09" with lines gnuplot> plot '/tmp/@20093.0' title "1115: 2000-03-06 17:23:09" with lines ^ can't read data file "/tmp/@20093.0" line 0: (No such file or directory) The second call to plot is done on an attempted mouse zoom. The same thing does work when fifos and inline are off, although in that case, plot is not called a 2nd time. Perhaps the 2nd call to plot with fifos is a gnuplot OS X oddity? Anyhow, mouse zooming works for me, so I'm happy! Cheers! Andrew On Wednesday, February 5, 2003, at 12:38 AM, Andrew Straw wrote: > It sounds good... However, I just did a CVS checkout, build, and > install and got the following: > > >>> import Gnuplot > Traceback (most recent call last): > File "<stdin>", line 1, in ? > File > "/Library/Frameworks/Python.framework/Versions/2.2/lib/python2.2/site- > packages/Gnuplot/__init__.py", line 177, in ? > from PlotItems import PlotItem, Func, File, Data, GridData > File > "/Library/Frameworks/Python.framework/Versions/2.2/lib/python2.2/site- > packages/Gnuplot/PlotItems.py", line 374, in ? > if gp.GnuplotOpts.support_fifo: > AttributeError: class GnuplotOpts has no attribute 'support_fifo' > > (I'll dig farther, but this is what I got to tonight.) > > Cheers! > Andrew > > On Tuesday, February 4, 2003, at 03:09 AM, Michael Haggerty wrote: > >> Andrew, >> >> Andrew Straw <and...@ad...> wrote: >>> First of all, a big thanks for contributing Gnuplot.py! I use it >>> all the time, as I'm sure countless others do. >> >> Thanks for the kind words. >> >>> Now a question -- is there a way to pipe data through a >>> Gnuplot.PlotItems.TempArrayFile by default? I ask because on my >>> machine (Mac OS X 10.2.3 with gnuplot 3.8h0), the 'set mouse' on the >>> x11 terminal only works when the data is available in such a file. >>> >>> I vaguely seem to recall another version of Gnuplot (I'm using 1.6) >>> did >>> this. (I remember lots of /tmp/@1234132.1 files.) I did find in the >>> gp_macosx.py file a "prefer_inline_data" variable, which I have set >>> to >>> 1, to no avail. >> >> That's funny; I thought that in version 1.6 temporary files were still >> the default for Mac OS X. >> >> In any case, in the current CVS version, FIFOs are used by default to >> send data to gnuplot on platforms that support them, including Mac OS >> X. In the CVS version, if you set the following options in >> gp_macos.py, then you should get temporary files by default: >> >> prefer_inline_data = 0 >> prefer_fifo_data = 1 >> >> You can obtain the CVS version from sourceforge; instructions are >> available one or two clicks from the Gnuplot.py project page, >> >> http://sourceforge.net/projects/gnuplot-py/ >> >> Also from the sourceforge project page you can find instructions how >> to join our mailing list, which is a good place to submit such >> questions and discussion. (I have forwarded a copy of this email to >> the mailing list.) >> >> Michael >> >> -- >> Michael Haggerty >> hag...@jp... >> > > > > ------------------------------------------------------- > This SF.NET email is sponsored by: > SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! > http://www.vasoftware.com > _______________________________________________ > Gnuplot-py-users mailing list > Gnu...@li... > https://lists.sourceforge.net/lists/listinfo/gnuplot-py-users > |
From: <kai...@t-...> - 2003-02-05 09:44:00
|
Andrew Straw wrote: > Sorry, my error report was premature -- deleting my old Gnuplot > module before re-installing fixes the problem. > > And I now have working mouse zoom! Unfortunately, I cannot get this > to work with fifos -- it seems the "file" is unlinked immediately > after gnuplot gets it the first time, and gnuplot wants it again when > trying to zoom. I assume that the mouse-zoom feature is implemented (under the hood) by calling the equivalent of "replot", which causes gnuplot to try to read the file again. The whole point of the FIFO code is that after Gnuplot.py has written the data once through the FIFO it can assume that gnuplot is done with the data and therefore the pseudo-file can be deleted. This, in turn, was a trick to get around the problem of otherwise not knowing when normal temporary files could be deleted. So I'm not surprised to hear that mouse-zoom doesn't work with FIFOs. If you can think of a strategy to get around this problem, please let me know. > The same thing does work when fifos and inline are off, although in > that case, plot is not called a 2nd time. Perhaps the 2nd call to > plot with fifos is a gnuplot OS X oddity? I hope so; otherwise the FIFO workaround for the temporary file problem isn't a good one! Michael -- Michael Haggerty mh...@al... |