Hi everyone, For those who are interested, I finally figured out what the gnuplot issue was, and I am stunned. It did not like using 1/4 and 1/2 as coefficients for the sin() functions. When I replaced them with 0.25 and 0.5 the plots work fine.
These would not work: $fnct_rms_halfwave="1/sqrt(pi) * sqrt(1/2*(pi-x*pi/180) + 1/4*sin(2*x*pi/180))"; $fnct_rms_fullwave="sqrt(2/pi) * sqrt(1/2*(pi-x*pi/180) + 1/4*sin(2*x*pi/180))"; These work fine: $fnct_rms_halfwave="1/sqrt(pi) * sqrt(0.5*(pi-x*pi/180) + 0.25*sin(2*x*pi/180))"; $fnct_rms_fullwave="sqrt(2/pi) * sqrt(0.5*(pi-x*pi/180) + 0.25*sin(2*x*pi/180))"; Just thought I'd share what I found in case anyone else runs across this problem. Although I don't know the "why", at least I know the "what" that caused it. If anyone knows the why, I'd love to know... BR, Gene L. Harding, PE Associate Professor of ECET Purdue University 574-520-4190 https://polytechnic.purdue.edu/south-bend/ -----Original Message----- From: LON-CAPA-users <lon-capa-users-boun...@mail.lon-capa.org> On Behalf Of raeb...@msu.edu Sent: Wednesday, April 8, 2020 8:27 AM To: Discussion list for LON-CAPA users <lon-capa-users@mail.lon-capa.org> Subject: Re: [LON-CAPA-users] gnuplot problem Hello Gene, > > "/home/httpd/perl/tmp/glhardin_purdue_1586186312_6039374_plot.data", > line 13: expecting ':' > Line 13 refers to the line number in the *_plot.data file, which would have been in /home/httpd/perl/tmp on the LON-CAPA server from where ITaP retrieved the error log information. Of course by now, nightly clean-up scripts will have removed the *_plot.data files from /home/httpd/perl/tmp, so you need to regenerate the error, and have ITaP provide you with the *_plot.data file which corresponds with the (new) error logged in the web server error logs. Once you have the *_plot.data file then you could also experiment with that outside LON-CAPA, by using it with a version of gnuplot installed locally on your own computer. If you wanted to experiment within the Linux environment, but don't use Linux as your primary OS, you could set up a Linux VM using VirtualBox (or similar), if you have the memory available to allocate to a VM, and install gnuplot there. Stuart Raeburn LON-CAPA Academic Consortium ________________________________________ From: LON-CAPA-users <lon-capa-users-boun...@mail.lon-capa.org> on behalf of Harding, Gene L <glhar...@purdue.edu> Sent: Tuesday, April 7, 2020 7:33 PM To: Discussion list for LON-CAPA users Subject: Re: [LON-CAPA-users] gnuplot problem Hi Stuart, Wouldn't taking the absolute value of the expression inside the sqrt function also work? I tried this, thinking it would remove the negative number problem, but it did not work: $fnct_rms_halfwave="1/sqrt(pi) * sqrt( abs(1/2*(pi-x*pi/180) + 1/4*sin(2*x*pi/180)) )"; I checked my gnuplot manual, and abs() appears to be the correct syntax for absolute value. Wouldn't that fix the negative number issue? These are the lines IT sent me from the error log: [Mon Apr 06 10:59:07 2020] [error] [client 128.210.56.177] "/home/httpd/perl/tmp/glhardin_purdue_1586185147_62400672_plot.data", line 13: expecting ':', referer: https://urldefense.com/v3/__https://loncapa.purdue.edu/priv/purdue/ecet_purdue/Topics/Power_Electronics/Thyristors/Calc_FiringAngle_using_Graphs_GeogebraInteractive_Multipart.problem__;!!HXCxUKc!gG78S6Qp2qDtQFc0NrfUKzJ4BmC7aOLMiBQ-9GfMLhgv8M9GqPiE0mgQfYzpVQ$ [Mon Apr 06 11:17:06 2020] [error] [client 128.210.56.177] "/home/httpd/perl/tmp/glhardin_purdue_1586186225_6003521_plot.data", line 13: expecting ':', referer: https://urldefense.com/v3/__https://loncapa.purdue.edu/priv/purdue/ecet_purdue/Topics/Power_Electronics/Thyristors/Calc_FiringAngle_using_Graphs_GeogebraInteractive_Multipart.problem__;!!HXCxUKc!gG78S6Qp2qDtQFc0NrfUKzJ4BmC7aOLMiBQ-9GfMLhgv8M9GqPiE0mgQfYzpVQ$ [Mon Apr 06 11:18:32 2020] [error] [client 128.210.56.177] "/home/httpd/perl/tmp/glhardin_purdue_1586186312_6039374_plot.data", line 13: expecting ':', referer: https://urldefense.com/v3/__https://loncapa.purdue.edu/priv/purdue/ecet_purdue/Topics/Power_Electronics/Thyristors/Calc_FiringAngle_using_Graphs_GeogebraInteractive_Multipart.problem__;!!HXCxUKc!gG78S6Qp2qDtQFc0NrfUKzJ4BmC7aOLMiBQ-9GfMLhgv8M9GqPiE0mgQfYzpVQ$ [Mon Apr 06 14:42:22 2020] [error] [client 128.210.56.177] "/home/httpd/perl/tmp/glhardin_purdue_1586198542_65331252_plot.data", line 13: expecting ':', referer: https://urldefense.com/v3/__https://loncapa.purdue.edu/priv/purdue/ecet_purdue/Topics/Power_Electronics/Thyristors/Calc_FiringAngle_using_Graphs_GeogebraInteractive_Multipart.problem__;!!HXCxUKc!gG78S6Qp2qDtQFc0NrfUKzJ4BmC7aOLMiBQ-9GfMLhgv8M9GqPiE0mgQfYzpVQ$ Do they make sense to you? Line 13 in the LON-CAPA code is part of a <startouttext /> statement, so I don't think that's the line the errors are referring to. I looked through the code again, and do not know where it might be expecting to see a ":" character, although I have done many versions of this during troubleshooting, and I wonder if this error came from a version when I was using the ternary operator to limit the plot interval. Gene L. Harding, PE Associate Professor of ECET Purdue University 574-520-4190 https://urldefense.com/v3/__https://polytechnic.purdue.edu/south-bend/__;!!HXCxUKc!gG78S6Qp2qDtQFc0NrfUKzJ4BmC7aOLMiBQ-9GfMLhgv8M9GqPiE0mgtNE-WWg$ -----Original Message----- From: LON-CAPA-users <lon-capa-users-boun...@mail.lon-capa.org> On Behalf Of raeb...@msu.edu Sent: Tuesday, April 7, 2020 8:50 AM To: Discussion list for LON-CAPA users <lon-capa-users@mail.lon-capa.org> Subject: Re: [LON-CAPA-users] gnuplot problem Gene, > > Two of them work fine, and the other two, which both have sine > expressions inside a sqrt function, do not. > My guess would be that gnuplot is encountering an issue because floating point arithmetic results in evaluation of the square root of a negative number (very close to zero), which should really be 0. In that case an alternative would be to explicitly evaluate the components of the function within a perl script block in LON-CAPA, and test that they will not result in a negative number before applying LON-CAPA's sqrt() function to the result (rom which you would then add the value to an array (@Y) of y values for the plot). The array (@X) will contain the corresponding x values, which you would have generating within the loop you use to generate the y values (in the @Y) array. You would then replace use of the <function></function> tag with <data>@X</data> <data>@Y</data> where @X and @Y are arrays containing x and y values for the points on the plot. See the: "Data Plot with Numerical Response" example in the problem templates. Stuart Raeburn LON-CAPA Academic Consortium ________________________________________ From: LON-CAPA-users <lon-capa-users-boun...@mail.lon-capa.org> on behalf of Harding, Gene L <glhar...@purdue.edu> Sent: Monday, April 6, 2020 7:35 PM To: Discussion list for LON-CAPA users Subject: Re: [LON-CAPA-users] gnuplot problem Thanks Stuart. I do not have access to those directories (as far as I can tell), so have reached out to some IT folks to see if they can help me. I realized later that, apparently, "dynamically generated plot" is just the alt tag for the gnuplot. I am baffled about what is going on. I have four functions for this problem. Two of them work fine, and the other two, which both have sine expressions inside a sqrt function, do not. I have tried using the ternary operator to zero out values outside of the axis limits. In all of the playing/troubleshooting I have done, it appears to not like the sqrt(bla bla sin(stuff)). I am not sure what version of gnuplot we are using. I think I am just going to do the problem using Geogebra and forget about gnuplot, but it would be nice to know what's going on with the gnuplot. Hopefully, my IT guys can help... BR, Gene L. Harding, PE Associate Professor of ECET Purdue University 574-520-4190 https://urldefense.com/v3/__https://polytechnic.purdue.edu/south-bend/__;!!HXCxUKc!jTjoyEwsimPmL2daj4jh1wXvO-ThMXDm58g_FiI6994pTe0-lVckSwwcNawAEA$ -----Original Message----- From: LON-CAPA-users <lon-capa-users-boun...@mail.lon-capa.org> On Behalf Of Raeburn, Stuart Sent: Monday, April 6, 2020 3:11 PM To: Discussion list for LON-CAPA users <lon-capa-users@mail.lon-capa.org> Subject: Re: [LON-CAPA-users] gnuplot problem Hello Gene, If the text: "dynamically generated plot" is displayed instead of an image, when using gnuplot, then it means gnuplot was unable to process the data file generated by LON-CAPA for your plot. That data file will have been written to /home/httpd/perl/tmp/ on your server's filesystem with a file name that includes your username, and ends _plot.data. and there will also be an error message in the web server's error log file, e.g., /var/log/httpd/ssl_error_log, with your problem listed as the referer. (You'd need to ask your system administrator to look in the log files). I was able to successfully generate plots for both functions: $fnct_rms_halfwave="1/$pi * &sqrt(1/2*($pi-x) + 1/4*sin(2*x))"; $fnct_rms_fullwave="&sqrt(2/$pi) * &sqrt(1/2*($pi-x) + 1/4*sin(2*x))"; in a problem in my Authoring Space (gnuplot version: 4.4.2). You might double check that there is not a typo in the problem XML, where you refer to the function, e.g., <function>$fnct_rms_halfwave</function> Otherwise, there could be an issue with other attributes defined for the plot, aside from the functions themselves (although in most cases, LON-CAPA replaces invalid values with valid ones, before creating the _plot.data file passed to gnuplot). Stuart Raeburn LON-CAPA Academic Consortium ________________________________________ From: LON-CAPA-users <lon-capa-users-boun...@mail.lon-capa.org> on behalf of Harding, Gene L <glhar...@purdue.edu> Sent: Monday, April 6, 2020 11:22 AM To: Discussion list for LON-CAPA users Subject: [LON-CAPA-users] gnuplot problem Hi everyone, I am having trouble generating a gnuplot using the following two functions in LON-CAPA: $fnct_rms_halfwave="1/$pi * &sqrt(1/2*($pi-x) + 1/4*sin(2*x))"; $fnct_rms_fullwave="&sqrt(2/$pi) * &sqrt(1/2*($pi-x) + 1/4*sin(2*x))"; Instead of rendering a plot, LON-CAPA just displays a message that says "dynamically generated plot". It plots fine in Geogebra. Does anyone know what's going on with the gnuplot? Best regards, Gene L. Harding, PE Associate Professor of ECET Purdue University 574-520-4190 https://urldefense.com/v3/__https://polytechnic.purdue.edu/south-bend/__;!!HXCxUKc!jTjoyEwsimPmL2daj4jh1wXvO-ThMXDm58g_FiI6994pTe0-lVckSwwcNawAEA$ _______________________________________________ LON-CAPA-users mailing list LON-CAPA-users@mail.lon-capa.org http://mail.lon-capa.org/mailman/listinfo/lon-capa-users _______________________________________________ LON-CAPA-users mailing list LON-CAPA-users@mail.lon-capa.org http://mail.lon-capa.org/mailman/listinfo/lon-capa-users _______________________________________________ LON-CAPA-users mailing list LON-CAPA-users@mail.lon-capa.org http://mail.lon-capa.org/mailman/listinfo/lon-capa-users
