You can subscribe to this list here.
2014 
_{Jan}

_{Feb}
(232) 
_{Mar}
(323) 
_{Apr}
(383) 
_{May}
(359) 
_{Jun}
(435) 
_{Jul}
(252) 
_{Aug}
(172) 
_{Sep}
(265) 
_{Oct}
(263) 
_{Nov}
(350) 
_{Dec}
(359) 

2015 
_{Jan}
(267) 
_{Feb}
(220) 
_{Mar}
(311) 
_{Apr}
(269) 
_{May}
(388) 
_{Jun}
(403) 
_{Jul}
(172) 
_{Aug}
(399) 
_{Sep}
(364) 
_{Oct}
(269) 
_{Nov}
(357) 
_{Dec}
(468) 
2016 
_{Jan}
(618) 
_{Feb}
(592) 
_{Mar}
(625) 
_{Apr}
(516) 
_{May}
(375) 
_{Jun}
(155) 
_{Jul}
(346) 
_{Aug}
(262) 
_{Sep}
(346) 
_{Oct}
(291) 
_{Nov}
(333) 
_{Dec}
(335) 
2017 
_{Jan}
(436) 
_{Feb}
(460) 
_{Mar}
(370) 
_{Apr}
(189) 
_{May}
(252) 
_{Jun}
(272) 
_{Jul}
(286) 
_{Aug}
(98) 
_{Sep}

_{Oct}

_{Nov}

_{Dec}

From: David Billinghurst <dbmaxima@gm...>  20170816 22:28:50

On 17/08/2017 03:08, Robert Dodier wrote: > My advice is to send technical, nonpersonal messages > to the mailing list, even if nobody expresses interest Agree. I found this exchange interesting, and had a play with it myself. 
From: Eric Barth <ejbarth@gm...>  20170816 20:15:10

Rather than "hold" as in matlab, put both functions in a single call to plot2d: plot2d([sin(2*x),sin(3*x)],[x,0,10],same_xy); Eric On Wed, Aug 16, 2017 at 3:57 PM Clinton Winant <clinton.winant@...> wrote: > Hi Kris, sorry for being unclear, my 2nd question was really asking how I > could invoke plot2d several times, w/o losing the previous plots, ie if I > wanted to replace > > plot(]sin(2*x),sin(3*x)],[x,0,1]) > > with two sequential calls to plot2d: > > plot2d(sin(2*x),[x,0,1]); > plot2d(sin(3*x),[x,0,1]); > > The way it is now the second plot clobbers the first. > > On Wed, Aug 16, 2017 at 12:07 PM, Kris Katterjohn <katterjohn@...> > wrote: > >> On Wed, Aug 16, 2017 at 10:46:55AM 0700, Clinton Winant wrote: >> > The following command works: >> > >> > >> plot2d([newpsi(Evalue[1],x),newpsi(Evalue[2],x),newpsi(Evalue[3],x),newpsi(Evalue[4],x)],[x,0,1]); >> > >> > but can the first [] be abbreviated somehow, for instance as >> > [newpsi(Evalue(k),x) ,k=1:4] >> >> This is just a list, so you can use things like makelist to construct >> it for you: >> >> makelist(newpsi(Evalue[k],x),k,1,4) >> >> See the docs for makelist for a description of the various ways it can >> be called. >> >> > Different question: >> > >> > What is the Maxima equivalent to Octave/Matlab "hold on" >> >> I don't know what "hold on" is. In general it's probably best to follow >> something like that with a description of what it actually does. >> >> Cheers, >> Kris Katterjohn >> > > >  > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Maximadiscuss mailing list > Maximadiscuss@... > https://lists.sourceforge.net/lists/listinfo/maximadiscuss > 
From: Clinton Winant <clinton.winant@gm...>  20170816 19:57:26

Hi Kris, sorry for being unclear, my 2nd question was really asking how I could invoke plot2d several times, w/o losing the previous plots, ie if I wanted to replace plot(]sin(2*x),sin(3*x)],[x,0,1]) with two sequential calls to plot2d: plot2d(sin(2*x),[x,0,1]); plot2d(sin(3*x),[x,0,1]); The way it is now the second plot clobbers the first. On Wed, Aug 16, 2017 at 12:07 PM, Kris Katterjohn <katterjohn@...> wrote: > On Wed, Aug 16, 2017 at 10:46:55AM 0700, Clinton Winant wrote: > > The following command works: > > > > plot2d([newpsi(Evalue[1],x),newpsi(Evalue[2],x),newpsi( > Evalue[3],x),newpsi(Evalue[4],x)],[x,0,1]); > > > > but can the first [] be abbreviated somehow, for instance as > > [newpsi(Evalue(k),x) ,k=1:4] > > This is just a list, so you can use things like makelist to construct > it for you: > > makelist(newpsi(Evalue[k],x),k,1,4) > > See the docs for makelist for a description of the various ways it can > be called. > > > Different question: > > > > What is the Maxima equivalent to Octave/Matlab "hold on" > > I don't know what "hold on" is. In general it's probably best to follow > something like that with a description of what it actually does. > > Cheers, > Kris Katterjohn > 
From: Martin Kraska <martin.kraska@th...>  20170816 19:53:56

I confirm that the latest nightly build solves the problem with access to gnuplot pdf files. Thanks a lot, Gunter. Best regards, Martin Kraska  Am 14. Aug 2017 um 9:37 schrieb Gunter Königsmann <gunter@...>: > Ok... ...as soon as my internet connection allows me to do so I will upload the > change that causes all files from draw to be closed. > Am 14. August 2017 08:40:40 MESZ schrieb Martin Kraska > <martin.kraska@...>: >> At gnuplot they say >>> A PDF file can contain many pages of output. How would gnuplot know whether you >>> are planning to add more pages or not? If you are done plotting and want to >>> close the file, tell the program either "unset output" or give a new "set >>> output". >> Maxima in principle seems to be aware of that point, because the handbook lists >>> terminal [Graphic option] >>> Default value: screen >>> Selects the terminal to be used by Gnuplot; possible values are: screen >>> (default), png, pngcairo, jpg, gif, eps, eps_color, epslatex, epslatex_ >>> standalone, svg, canvas, dumb, dumb_file, pdf, pdfcairo , wxt, animated_gif, >>> multipage_pdfcairo, multipage_pdf , multipage_eps, multipage_eps_color, >>> and aquaterm. >> The multipage terminals aren't genuine gnuplot terminals but some Maxima >> special. In this case, whenever a nonmultipage terminal is specified, the last >> command in the gnuplot command file could safely be "set output" to close the >> file. >> Martin Kraska >>  Am 13. Aug 2017 um 22:09 schrieb Martin Kraska >> <martin.kraska@...>: >>> Ok, I filed a bug report: https://sourceforge.net/p/gnuplot/bugs/1955/ >>> I use Maxima with SBCL under windows and the unicode support is fine. >>> Best regards, Martin Kraska >>>  Am 13. Aug 2017 um 19:46 schrieb Gunter Königsmann <gunter@...>: >>>> Fixing it on the gnuplot side would be better: this way the fix would work for >>>> all projects using the program. >>>> With SciTe I have no experience. >>>> What I know is that Maxima by default is set up to support Unicode characters in >>>> variable names if it either was compiled using sbcl, isn't running on windows >>>> or both. With all other lisps the user currently has either to find out how to >>>> teach the lisp to support Unicode or how to tell it to completely skip the test >>>> if the input is encoded with Unicode or with the local codepage. Maxima will >>>> accept Unicode Symbol names in both cases. >>>> Kind regards, >>>> Gunter. >  > Diese Nachricht wurde von meinem AndroidGerät mit K9 Mail gesendet. 
From: Kris Katterjohn <katterjohn@gm...>  20170816 19:08:07

On Wed, Aug 16, 2017 at 10:46:55AM 0700, Clinton Winant wrote: > The following command works: > > plot2d([newpsi(Evalue[1],x),newpsi(Evalue[2],x),newpsi(Evalue[3],x),newpsi(Evalue[4],x)],[x,0,1]); > > but can the first [] be abbreviated somehow, for instance as > [newpsi(Evalue(k),x) ,k=1:4] This is just a list, so you can use things like makelist to construct it for you: makelist(newpsi(Evalue[k],x),k,1,4) See the docs for makelist for a description of the various ways it can be called. > Different question: > > What is the Maxima equivalent to Octave/Matlab "hold on" I don't know what "hold on" is. In general it's probably best to follow something like that with a description of what it actually does. Cheers, Kris Katterjohn 
From: Clinton Winant <clinton.winant@gm...>  20170816 17:47:03

The following command works: plot2d([newpsi(Evalue[1],x),newpsi(Evalue[2],x),newpsi(Evalue[3],x),newpsi(Evalue[4],x)],[x,0,1]); but can the first [] be abbreviated somehow, for instance as [newpsi(Evalue(k),x) ,k=1:4] Different question: What is the Maxima equivalent to Octave/Matlab "hold on" 
From: Robert Dodier <robert.dodier@gm...>  20170816 17:08:14

On Wed, Aug 16, 2017 at 1:38 AM, Soegtrop, Michael <michael.soegtrop@...> wrote: > The downside of this is that the nullspace computation with solve > is a bit shaky, I wanted to replace this with a SVD. There is a numerical SVD function in Maxima's lapack package, named dgesvd. I don't know of a symbolic SVD; such a thing does seem useful. > If someone else is interested in this, please let me know. My advice is to send technical, nonpersonal messages to the mailing list, even if nobody expresses interest, since it makes the messages available in the mailing list archive for someone to find later. For my own part, I find it interesting to hear what others are working on, even if I am not participating. best Robert Dodier 
From: Soegtrop, Michael <michael.soegtrop@in...>  20170816 08:39:06

Dear Clinton, I sent you a private mail which shows how to get several eigenvalues and also plot eigenfunctions. It is a bit more complicated in that it has in between boundary and continuity conditions and a piecewise solution. The downside of this is that the nullspace computation with solve is a bit shaky, I wanted to replace this with a SVD. If someone else is interested in this, please let me know. Best regards, Michael Intel Deutschland GmbH Registered Address: Am Campeon 1012, 85579 Neubiberg, Germany Tel: +49 89 99 88530, http://www.intel.de Managing Directors: Christin Eisenschmid, Christian Lamprechter Chairperson of the Supervisory Board: Nicole Lau Registered Office: Munich Commercial Register: Amtsgericht Muenchen HRB 186928 
From: Gunter Königsmann <gunter@pe...>  20170815 19:36:32

makelist(value (i),i,1,5); Kind regards, Gunter. Am 15. August 2017 20:19:27 MESZ schrieb Clinton Winant <clinton.winant@...>: >I list the fist five eigenvalues of a problem with the following >script: > >dispersion:cos(k)=sech(k); >value(m,k):= block([], >if evenp(m) >then find_root(dispersion,k,m*%pi,(2*m+1)*%pi/2) >else find_root(dispersion,k,(2*m+1)*%pi/2,(m+1)*%pi) >); >for m:1 thru 5 do disp(value(m,k)); > >(1) I want to modify the script so that the eigenvalues are in a list, >say >Evalue(1:n). > Right now, If I enter value; I just get "value" returned > >(2) I want to plot the corresponding modes: >mode(n,x)=sin(Evalue(n)*x) over the domain x=[0,1] > >I know to use plot2d as in: >plot2d([cos(k),sech(k)],[k,0,1]) >but dont see how to extend this to the Evectors ?? > >Thanks  Diese Nachricht wurde von meinem AndroidGerät mit K9 Mail gesendet. 
From: Clinton Winant <clinton.winant@gm...>  20170815 18:19:35

I list the fist five eigenvalues of a problem with the following script: dispersion:cos(k)=sech(k); value(m,k):= block([], if evenp(m) then find_root(dispersion,k,m*%pi,(2*m+1)*%pi/2) else find_root(dispersion,k,(2*m+1)*%pi/2,(m+1)*%pi) ); for m:1 thru 5 do disp(value(m,k)); (1) I want to modify the script so that the eigenvalues are in a list, say Evalue(1:n). Right now, If I enter value; I just get "value" returned (2) I want to plot the corresponding modes: mode(n,x)=sin(Evalue(n)*x) over the domain x=[0,1] I know to use plot2d as in: plot2d([cos(k),sech(k)],[k,0,1]) but dont see how to extend this to the Evectors ?? Thanks 
From: Martin Kraska <martin.kraska@th...>  20170815 16:38:31

Thanks a lot. I'll give it a try (never used nightly builds of Maxima) I guess that the gnuplot files are closed only for the nonmultipage terminals... Best regards, Martin Kraska  Am 14. Aug 2017 um 9:37 schrieb Gunter Königsmann <gunter@...>: > Ok... ...as soon as my internet connection allows me to do so I will upload the > change that causes all files from draw to be closed. > Am 14. August 2017 08:40:40 MESZ schrieb Martin Kraska > <martin.kraska@...>: >> At gnuplot they say >>> A PDF file can contain many pages of output. How would gnuplot know whether you >>> are planning to add more pages or not? If you are done plotting and want to >>> close the file, tell the program either "unset output" or give a new "set >>> output". >> Maxima in principle seems to be aware of that point, because the handbook lists >>> terminal [Graphic option] >>> Default value: screen >>> Selects the terminal to be used by Gnuplot; possible values are: screen >>> (default), png, pngcairo, jpg, gif, eps, eps_color, epslatex, epslatex_ >>> standalone, svg, canvas, dumb, dumb_file, pdf, pdfcairo , wxt, animated_gif, >>> multipage_pdfcairo, multipage_pdf , multipage_eps, multipage_eps_color, >>> and aquaterm. >> The multipage terminals aren't genuine gnuplot terminals but some Maxima >> special. In this case, whenever a nonmultipage terminal is specified, the last >> command in the gnuplot command file could safely be "set output" to close the >> file. >> Martin Kraska >>  Am 13. Aug 2017 um 22:09 schrieb Martin Kraska >> <martin.kraska@...>: >>> Ok, I filed a bug report: https://sourceforge.net/p/gnuplot/bugs/1955/ >>> I use Maxima with SBCL under windows and the unicode support is fine. >>> Best regards, Martin Kraska >>>  Am 13. Aug 2017 um 19:46 schrieb Gunter Königsmann <gunter@...>: >>>> Fixing it on the gnuplot side would be better: this way the fix would work for >>>> all projects using the program. >>>> With SciTe I have no experience. >>>> What I know is that Maxima by default is set up to support Unicode characters in >>>> variable names if it either was compiled using sbcl, isn't running on windows >>>> or both. With all other lisps the user currently has either to find out how to >>>> teach the lisp to support Unicode or how to tell it to completely skip the test >>>> if the input is encoded with Unicode or with the local codepage. Maxima will >>>> accept Unicode Symbol names in both cases. >>>> Kind regards, >>>> Gunter. >  > Diese Nachricht wurde von meinem AndroidGerät mit K9 Mail gesendet. 
From: Raymond Toy <toy.raymond@gm...>  20170815 14:35:55

The failure is from sbcl, compiling ifactor.lisp: ; compiling (DEFUN PRIMEPPROB ...)fatal error encountered in SBCL pid 16971(tid 46912496329856): no transport function for object 0x0e59800f (widetag 0xf3) On Tue, Aug 15, 2017 at 6:45 AM, Gunter Königsmann <gunter@...> wrote: > Currently my internet connection is too flaky to download the build log. > But today I got a mail about a build failure that wasn't there before. Hope > it wasn't my commit. Forwarding the mail from the build daemon. > >  > *Von:* Launchpad Buildd System <noreply@...> > *Gesendet:* 15. August 2017 11:18:03 MESZ > *An:* PeterPall <gunter@...> > *Betreff:* [Build #13249571] amd64 build of maxima 5.40.02~201708141101~ubuntu15.04.1 > in ubuntu vivid RELEASE [~maximadevelopers/ubuntu/maximanightly] > > > * Source Package: maxima > * Version: 5.40.02~201708141101~ubuntu15.04.1 > * Architecture: amd64 > * Archive: ~maximadevelopers/ubuntu/maximanightly > * Component: main > * State: Failed to build > * Duration: 12 minutes > * Build Log: https://launchpad.net/~maximadevelopers/+archive/ubuntu/maximanightly/+build/13249571/+files/buildlog_ubuntuvividamd64.maxima_5.40.02~201708141101~ubuntu15.04.1_BUILDING.txt.gz > * Builder: https://launchpad.net/builders/lcy0112 > * Source: not available > > > > If you want further information about this situation, feel free to > contact a member of the Launchpad Buildd Administrators team. > >  > amd64 build of maxima 5.40.02~201708141101~ubuntu15.04.1 in ubuntu vivid RELEASE > https://launchpad.net/~maximadevelopers/+archive/ubuntu/maximanightly/+build/13249571 > > You are receiving this email because your team The maxima developers is > the owner of this archive. > > >  > Diese Nachricht wurde von meinem AndroidGerät mit K9 Mail gesendet. >  >  > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Maximadiscuss mailing list > Maximadiscuss@... > https://lists.sourceforge.net/lists/listinfo/maximadiscuss > >  Ray 
From: Gunter Königsmann <gunter@pe...>  20170815 13:46:00

Currently my internet connection is too flaky to download the build log. But today I got a mail about a build failure that wasn't there before. Hope it wasn't my commit. Forwarding the mail from the build daemon.  Ursprüngliche Nachricht  Von: Launchpad Buildd System <noreply@...> Gesendet: 15. August 2017 11:18:03 MESZ An: PeterPall <gunter@...> Betreff: [Build #13249571] amd64 build of maxima 5.40.02~201708141101~ubuntu15.04.1 in ubuntu vivid RELEASE [~maximadevelopers/ubuntu/maximanightly] * Source Package: maxima * Version: 5.40.02~201708141101~ubuntu15.04.1 * Architecture: amd64 * Archive: ~maximadevelopers/ubuntu/maximanightly * Component: main * State: Failed to build * Duration: 12 minutes * Build Log: https://launchpad.net/~maximadevelopers/+archive/ubuntu/maximanightly/+build/13249571/+files/buildlog_ubuntuvividamd64.maxima_5.40.02~201708141101~ubuntu15.04.1_BUILDING.txt.gz * Builder: https://launchpad.net/builders/lcy0112 * Source: not available If you want further information about this situation, feel free to contact a member of the Launchpad Buildd Administrators team.  amd64 build of maxima 5.40.02~201708141101~ubuntu15.04.1 in ubuntu vivid RELEASE https://launchpad.net/~maximadevelopers/+archive/ubuntu/maximanightly/+build/13249571 You are receiving this email because your team The maxima developers is the owner of this archive.  Diese Nachricht wurde von meinem AndroidGerät mit K9 Mail gesendet. 
From: Clinton Winant <clinton.winant@gm...>  20170814 23:23:49

Michael The hint you sent this morning helped me find a solution to a slightly simpler, better known problem,corresponding to alpha=0, as listed below: /* Find eigenvalues/vectors of the one dimensional biharmonic eqn */ /* d^4 psi/dx^4=lambda psi(x), psi=d psi/dx at x=0,1 */ /* Since we are intersted in real solutions only, a first try is */ psi:sin(k*x)+C*cos(k*x)+E*sinh(k*x)+F*cosh(k*x); /* psi will be detrmined to within an arbitrary constant */ /* Use 3 BC to determine C, E and F */ dpsidx:diff(psi,x,1); bc2at0:ev(dpsidx,x=0); bc1at1:ev(psi,x=1); bc2at1:ev(dpsidx,x=1); sol:solve([bc2at0,bc1at1,bc2at1],[C,E,F]); C:rhs(sol[1][1]); E:rhs(sol[1][2]); F:rhs(sol[1][3]); /* rewrite psi, and evaluate the fourth BC to get equation for k */ psi:ev(psi); dpsidx:diff(psi,x,1); bc1at0:ev(psi,x=0); bc2at0:ev(dpsidx,x=0); bc1at1:ev(psi,x=1); factor(bc1at1); bc2at1:ev(dpsidx,x=1); factor(bc2at1); /* rewrite dispersion relation as */ disp:cos(k)=sech(k); l:find_root(disp,k,0,%pi/2); /* following lines: first 3 eigenvalues */ /* 4.712389 */ /* 7.853982 */ /* 10.99557 */ /*rewrite psi, and check all bc are met*/ newpsi:(cos(k)*sinh(k)+cosh(k)*sin(k))*(sin(k*x)sinh(k*x))+(cos(k)*cosh(k)+sinh(k)*sin(k)1)*cosh(k*x)+(cos(k)*cosh(k)sinh(k)*sin(k)1)*cos(k*x); dnewpsidx:diff(newpsi,x,1); bc1at0:ev(newpsi,x=0); bc2at0:ev(dnewpsidx,x=0); bc1at1:ev(newpsi,x=1); factor(bc1at1);/* =0 */ bc2at1:ev(dnewpsidx,x=1); factor(bc2at1);/* =0 */ You will immediately see what a beginner I am in maxima: in several places where I check to see whether the BC are satisfied, I had to complete the calculation by hand. Also there must be a way to wrap the find_root call in a do loop, so several are found (here after the first 2 or 3, it is just lambda=(2n+1)pi/2, n=1,2.. n=0 is NOT ok). Anu suggestion on how to complete the BC evaluation to yield 0 in the last line and two line before that would be great. Could anyone suggest also how to plot the first few eigenvectors? I have also solved the problem I originally wrote (alpha not=0), but don't think anyone really cares to see both. Maxima is AWESOME On Mon, Aug 14, 2017 at 5:05 AM, Soegtrop, Michael < michael.soegtrop@...> wrote: > Dear Clinton, > > > > I took that as my lunch time Sudoku ;) > > > > Attached please find a .wxm file which gets you to the determinant. It > needs some knowledge of the domain of alpha and lambda (real or complex) to > 0 it. The solutions are computed with deslove and manual factorization of > the denominator of the laplace transform. I don’t know why Maxima doesn’t > do this automatically. The simplifier goes wild at the x=1 boundary > conditions, so I had to switch it off temporarily. I will report a bug on > this. > > > > After you determined alpha and lambda such that the determinant gets 0, > you need to compute the nullspace of CM. If you find an analytical solution > for alpha and lambda, you can use nullspace for this. If your solution is > numeric, you will need a SVD factorizer to compute the nullspace (I can > provide one for Maxima). > > > > This nullspace of CM gives the solution for c[i] and your final solution > function. > > > > If you found non trivial (lambda !=0) analytical values for alpha and > lambda, I would like to know what these are. > > > > Best regards, > > > > Michael > > > > > > Intel Deutschland GmbH > Registered Address: Am Campeon 1012, 85579 Neubiberg, Germany > Tel: +49 89 99 88530, http://www.intel.de > Managing Directors: Christin Eisenschmid, Christian Lamprechter > Chairperson of the Supervisory Board: Nicole Lau > Registered Office: Munich > Commercial Register: Amtsgericht Muenchen HRB 186928 > 
From: Soegtrop, Michael <michael.soegtrop@in...>  20170814 16:36:57

Dear Maxima Users, Is there a way to change the summation order of sums, e.g. convert sum( sum(x[i,j], i,1,N), j,1,N); such that the sum over i is outside and the sum over j is inside? Also are there functions to distribute and contract sums over sums of terms? Best regards, Michael Intel Deutschland GmbH Registered Address: Am Campeon 1012, 85579 Neubiberg, Germany Tel: +49 89 99 88530, http://www.intel.de Managing Directors: Christin Eisenschmid, Christian Lamprechter Chairperson of the Supervisory Board: Nicole Lau Registered Office: Munich Commercial Register: Amtsgericht Muenchen HRB 186928 
From: Raymond Toy <toy.raymond@gm...>  20170814 15:02:47

I wish I knew because I really liked using gmane to read my mailing lists. There seems to be no contact information anywhere at gmane.org. Maybe the best way is to contact Yomura who took over gmane from Lars? On Sat, Aug 12, 2017 at 12:07 PM, Robert Dodier <robert.dodier@...> wrote: > Hi, gmane.comp.mathematics.maxima.general hasn't been updated in > several days and one possibility is that Sourceforge has recently > required all mailing list users to confirm their subscriptions and the > mailing list robot did not. (Neither did robert.dodier@... so > I'm not one to blame anybody or anything.) I've tried to contact the > current management at Gmane but the project appears to be running on > fumes at this point. > > So, this is a long shot, but: does anyone know the email address of > the Gmane mailing list robot? I am thinking I can put it back on the > list of subscribers by hand. > > I really like NNTP for a few reasons so I have some motivation to get > this working ... Thanks for your help. > > Robert Dodier > >  >  > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Maximadiscuss mailing list > Maximadiscuss@... > https://lists.sourceforge.net/lists/listinfo/maximadiscuss >  Ray 
From: Soegtrop, Michael <michael.soegtrop@in...>  20170814 12:05:41

Dear Clinton, I took that as my lunch time Sudoku ;) Attached please find a .wxm file which gets you to the determinant. It needs some knowledge of the domain of alpha and lambda (real or complex) to 0 it. The solutions are computed with deslove and manual factorization of the denominator of the laplace transform. I don’t know why Maxima doesn’t do this automatically. The simplifier goes wild at the x=1 boundary conditions, so I had to switch it off temporarily. I will report a bug on this. After you determined alpha and lambda such that the determinant gets 0, you need to compute the nullspace of CM. If you find an analytical solution for alpha and lambda, you can use nullspace for this. If your solution is numeric, you will need a SVD factorizer to compute the nullspace (I can provide one for Maxima). This nullspace of CM gives the solution for c[i] and your final solution function. If you found non trivial (lambda !=0) analytical values for alpha and lambda, I would like to know what these are. Best regards, Michael Intel Deutschland GmbH Registered Address: Am Campeon 1012, 85579 Neubiberg, Germany Tel: +49 89 99 88530, http://www.intel.de Managing Directors: Christin Eisenschmid, Christian Lamprechter Chairperson of the Supervisory Board: Nicole Lau Registered Office: Munich Commercial Register: Amtsgericht Muenchen HRB 186928 
From: Gunter Königsmann <gunter@pe...>  20170814 10:40:45

Ok... ...as soon as my internet connection allows me to do so I will upload the change that causes all files from draw to be closed. Am 14. August 2017 08:40:40 MESZ schrieb Martin Kraska <martin.kraska@...>: >At gnuplot they say > > > >A PDF file can contain many pages of output. How would gnuplot know >whether you are planning to add more pages or not? If you are done >plotting and want to close the file, tell the program either "unset >output" or give a new "set output". > > > >Maxima in principle seems to be aware of that point, because the >handbook lists > >BQ_BEGIN > >terminal [Graphic option] >Default value: screen >Selects the terminal to be used by Gnuplot; possible values are: screen > >(default), png, pngcairo, jpg, gif, eps, eps_color, epslatex, epslatex_ > >standalone, svg, canvas, dumb, dumb_file, pdf, pdfcairo , wxt, >animated_gif, >multipage_pdfcairo, multipage_pdf , multipage_eps, multipage_eps_color, > >and aquaterm. > >BQ_END > >The multipage terminals aren't genuine gnuplot terminals but some >Maxima special. In this case, whenever a nonmultipage terminal is >specified, the last command in the gnuplot command file could safely be >"set output" to close the file. > >Martin Kraska > > > Am 13. Aug 2017 um 22:09 schrieb Martin Kraska ><martin.kraska@...>: > >BQ_BEGIN > >Ok, I filed a bug report: https://sourceforge.net/p/gnuplot/bugs/1955/ > >I use Maxima with SBCL under windows and the unicode support is fine. > >Best regards, Martin Kraska > > Am 13. Aug 2017 um 19:46 schrieb Gunter Königsmann ><gunter@...>: > >BQ_BEGIN >Fixing it on the gnuplot side would be better: this way the fix would >work for all projects using the program. > >With SciTe I have no experience. > >What I know is that Maxima by default is set up to support Unicode >characters in variable names if it either was compiled using sbcl, >isn't running on windows or both. With all other lisps the user >currently has either to find out how to teach the lisp to support >Unicode or how to tell it to completely skip the test if the input is >encoded with Unicode or with the local codepage. Maxima will accept >Unicode Symbol names in both cases. > >Kind regards, > >Gunter. >BQ_END > > >BQ_END  Diese Nachricht wurde von meinem AndroidGerät mit K9 Mail gesendet. 
From: Daniel Volinski <danielvolinski@ya...>  20170814 09:44:37

Hi Daniel I have attached some wxMaxima files with some ideas I used a while ago. Let me know if this helps or you have any questions. Daniel Volinski El Domingo 13 de agosto de 2017 6:20, Daniel Muller <muller@...> escribió: Hi allI'm very new with maxima and was trying to build up a code that could deal with frame basis. I look into a previous thread in Maxima's mailing list and found a code that obtains the divergence of a vector field in frame basis. I'm trying to generalize this previous code such that the it also calculates the curl of a vector field with the following code but was upset. I already received some help, but still the code is does not work. Can anyone help me? Best, Daniel/* test*/load(itensor);load(ctensor);depends([V,ifg,ifr,ifri],ct_coords)$iframe_flag:true;cframe_flag:true;ct_coordsys(spherical)$cmetric();fri;lfg;lg:trigsimp(lg);ishow(covdiff(V([],[j]),j))$ishow(ev(%,icc2,ifc2,ifc1,ifb))$eqn:eq=%$el1:ishow(contract('levi_civita([k,l,i],[])*ifg([],[j,l])*covdiff(V([],[i]),j)))$ ishow(canform(contract(expand(applyb1(el1,lc_l,lc_u)))))$eqn1:ishow(rotV([k],[])=ev(%,icc2,ifc2,ifc1,ifb))$ceqn1:ic_convert(rename(expand(eqn1)))$ceqn1:subst(ifri,nounify(ifri),ceqn1)$ceqn:ic_convert(rename(expand(eqn)))$ceqn:subst(ifri,nounify(ifri),ceqn)$ifg:matrix([1,0,0],[0,1,0],[0,0,1]);ifr:matrix([1,0,0],[0,1/r,0],[0,0,1/(r*sin(theta))]);ifri:matrix([1,0,0],[0,r,0],[0,0,r*sin(theta)]);rotV:[0,0,0];ceqn1,'ifr,'ifg;ceqn,'ifr,'ifg;for i thru dim do for j thru dim do for k thru dim do levi_civita[i,j,k]:0;levi_civita[1,2,3]:levi_civita[2,3,1]:levi_civita[3,1,2]:1;levi_civita[1,3,2]:levi_civita[3,2,1]:levi_civita[2,1,3]:1;rotV,levi_civita; Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot_______________________________________________ Maximadiscuss mailing list Maximadiscuss@... https://lists.sourceforge.net/lists/listinfo/maximadiscuss 
From: Soegtrop, Michael <michael.soegtrop@in...>  20170814 08:33:51

Dear Clinton, I have no solution to your problem, but a few general hints. For a 4th order homogeneous differential equation you typically need 4 independent “Ansatz” functions for the same values of k, alpha and lambda. I could find only two obvious ones: (%i1) NM:'diff(psi,x,4)+2*alpha*'diff(psi,x,2)=λ*psi; (NM)'diff(psi,x,4)+2*alpha*('diff(psi,x,2))=psi*λ (%i4) psi:exp(%i*k*x)$ radcan(ev(NM,eval,nouns))$ %/psi; (%o4) k^42*alpha*k^2=λ (%i7) psi:exp(%i*k*x)$ radcan(ev(NM,eval,nouns))$ %/psi; (%o7) k^42*alpha*k^2=λ You need 2 more that lead to the same equations. Or 4 different functions which all together lead to a different equation than above, but the same for all 4. Without the second order term, these would also work: (%i10) psi:exp(k*x)$ radcan(ev(NM,eval,nouns))$ %/psi; (%o10) k^4+2*alpha*k^2=λ (%i13) psi:exp(k*x)$ radcan(ev(NM,eval,nouns))$ %/psi; (%o13) k^4+2*alpha*k^2=λ but with the second order term they results in different equations. Btw.: the real parts of these 4 functions can be combined to sin, cos, sinh, cosh. If you are interested in real solutions, it makes sense to use real ansatz functions. If you have the 4 ansatz functions, the solution is a linear combination of these. The 4 boundary conditions lead to 4 linear equations for the constant coefficients with the ansatz functions as “variables”. You build a coefficient matrix of the 4 boundary equations for the 4 constant coefficients (using coeffmatrix). The entries of this matrix will be some combinations of derivatives of your ansatz functions. Since your differential equation is homogeneous, the equations only have a solution if the determinant of this matrix is 0. The parameter k is determined such that this is the case. This typically results in an equation which cannot be solved symbolically. You can plot the determinant over k and use a numeric solver at the roots you see (typically infinitely many). With k determined in this way, you end up with a numeric matrix for the coefficients. The solution is the nullspace of this matrix (the eigenvector of the 0 eigenvalue). The result is a linear combination of your 4 ansatz functions with numeric coefficients, so something semi symbolic. One more note: you might get away with only 2 Ansatz functions, but then you can only fulfill 2 border conditions. I can send you a wxMaxima file with the solution of a pole oscillation. It is also 4th order in the space domain, but doesn’t have the 2nd order term. This example has the additional complexity of a boundary conditions in the middle, so that I have a piecewise solution with continuity equations. The maxima solutions uses arrays effectively for this. Best regards, Michael Intel Deutschland GmbH Registered Address: Am Campeon 1012, 85579 Neubiberg, Germany Tel: +49 89 99 88530, http://www.intel.de Managing Directors: Christin Eisenschmid, Christian Lamprechter Chairperson of the Supervisory Board: Nicole Lau Registered Office: Munich Commercial Register: Amtsgericht Muenchen HRB 186928 
From: Stefano Milani <milani@en...>  20170814 07:25:59

Dear member, is possible to create an indipendent executable tool for my project directly with maxima, like octave and matlab? I would like to create a project to share with my colleagues that know know maxima. Best regards  Stefano Milani 
From: Martin Kraska <martin.kraska@th...>  20170814 06:40:59

At gnuplot they say A PDF file can contain many pages of output. How would gnuplot know whether you are planning to add more pages or not? If you are done plotting and want to close the file, tell the program either "unset output" or give a new "set output". Maxima in principle seems to be aware of that point, because the handbook lists BQ_BEGIN terminal [Graphic option] Default value: screen Selects the terminal to be used by Gnuplot; possible values are: screen (default), png, pngcairo, jpg, gif, eps, eps_color, epslatex, epslatex_ standalone, svg, canvas, dumb, dumb_file, pdf, pdfcairo , wxt, animated_gif, multipage_pdfcairo, multipage_pdf , multipage_eps, multipage_eps_color, and aquaterm. BQ_END The multipage terminals aren't genuine gnuplot terminals but some Maxima special. In this case, whenever a nonmultipage terminal is specified, the last command in the gnuplot command file could safely be "set output" to close the file. Martin Kraska  Am 13. Aug 2017 um 22:09 schrieb Martin Kraska <martin.kraska@...>: BQ_BEGIN Ok, I filed a bug report: https://sourceforge.net/p/gnuplot/bugs/1955/ I use Maxima with SBCL under windows and the unicode support is fine. Best regards, Martin Kraska  Am 13. Aug 2017 um 19:46 schrieb Gunter Königsmann <gunter@...>: BQ_BEGIN Fixing it on the gnuplot side would be better: this way the fix would work for all projects using the program. With SciTe I have no experience. What I know is that Maxima by default is set up to support Unicode characters in variable names if it either was compiled using sbcl, isn't running on windows or both. With all other lisps the user currently has either to find out how to teach the lisp to support Unicode or how to tell it to completely skip the test if the input is encoded with Unicode or with the local codepage. Maxima will accept Unicode Symbol names in both cases. Kind regards, Gunter. BQ_END BQ_END 
From: Martin Kraska <martin.kraska@th...>  20170813 20:09:34

Ok, I filed a bug report: https://sourceforge.net/p/gnuplot/bugs/1955/ I use Maxima with SBCL under windows and the unicode support is fine. Best regards, Martin Kraska  Am 13. Aug 2017 um 19:46 schrieb Gunter Königsmann <gunter@...>: > Fixing it on the gnuplot side would be better: this way the fix would work for > all projects using the program. > With SciTe I have no experience. > What I know is that Maxima by default is set up to support Unicode characters in > variable names if it either was compiled using sbcl, isn't running on windows > or both. With all other lisps the user currently has either to find out how to > teach the lisp to support Unicode or how to tell it to completely skip the test > if the input is encoded with Unicode or with the local codepage. Maxima will > accept Unicode Symbol names in both cases. > Kind regards, > Gunter. 
From: Gunter Königsmann <gunter@pe...>  20170813 17:46:40

Fixing it on the gnuplot side would be better: this way the fix would work for all projects using the program. With SciTe I have no experience. What I know is that Maxima by default is set up to support Unicode characters in variable names if it either was compiled using sbcl, isn't running on windows or both. With all other lisps the user currently has either to find out how to teach the lisp to support Unicode or how to tell it to completely skip the test if the input is encoded with Unicode or with the local codepage. Maxima will accept Unicode Symbol names in both cases. Kind regards, Gunter. Am 13. August 2017 15:32:04 MESZ schrieb Martin Kraska <martin.kraska@...>: >Hello, > >thank you for your response. In fact I think it is a file system issue. >I operate maxima via a socket connection which is utf8 encoded. If I >visit the gnuplot control file in SciTe, some characters are displayed >wrongly, however, all annotations with special characters (e.g. greek) >are handled correctly in the plots (tried pngcairo, svg, pdfcairo). The >output file, however, appears with the mangled name in the file system. > >I worked around this problem by sending a temporary random file name >without dangerous characters and then copy the result to the target >file name. >This works nicely for png and svg but failed for pdf. > >It seems that gnuplot doesn't close the pdf file correctly, you >explicitly have to send "set output" as a last command in order to get >the file unlocked under >windows. Therefore, for png the copy step of the workaround doesn't >work. > >This could be mitigated from maxima side (by resetting the output) or >from gnuplot side by closing the output after processing the input >file. > >Best regards, Martin Kraska > > Am 13. Aug 2017 um 9:38 schrieb Gunter Königsmann >gunter@...: > >>> Hmm, are you sure that filenames with UTF8 characters are displayed >>> corectly in general? To judge by the comments here: >>> >https://superuser.com/questions/157207/onwindows7dirortreecantshowunicodecharactersevenstartingcmdwith >>> it is a little bit tricky. A web search for "win7 utf8 characters >in >>> file names" seems to turn up some problems. >>> >>> I think draw2d will create a file named maxoutnnnn.gnuplot where >nnnn >>> is a process ID or something like that. If you look at that with a >>> text editor, do you see the file name correctly? If the file name >>> seems incorrect, can you verify that the bytes are correct somehow? >> >> The file will be in the directory indicated by maxima when typing >> >> maxima_tempdir; >> >> Another test would be: >> >> wxdraw2d(key="▶±ħ↝üØλΕφ",explicit(sin(x),x,1,10)); >> >> If the key is displayed correctly the file maxima passes to gnuplot >> definitively contains unicode characters (no single codepage contains >> greek, german and mathematical symbols). >> ...in this case we can hope that gnuplot is the problem: Normally the >> gnuplot programmers are quite fast and willing to help. >> >> Kind regards, >> >> Gunter.  Diese Nachricht wurde von meinem AndroidGerät mit K9 Mail gesendet. 
From: Martin Kraska <martin.kraska@th...>  20170813 13:32:21

Hello, thank you for your response. In fact I think it is a file system issue. I operate maxima via a socket connection which is utf8 encoded. If I visit the gnuplot control file in SciTe, some characters are displayed wrongly, however, all annotations with special characters (e.g. greek) are handled correctly in the plots (tried pngcairo, svg, pdfcairo). The output file, however, appears with the mangled name in the file system. I worked around this problem by sending a temporary random file name without dangerous characters and then copy the result to the target file name. This works nicely for png and svg but failed for pdf. It seems that gnuplot doesn't close the pdf file correctly, you explicitly have to send "set output" as a last command in order to get the file unlocked under windows. Therefore, for png the copy step of the workaround doesn't work. This could be mitigated from maxima side (by resetting the output) or from gnuplot side by closing the output after processing the input file. Best regards, Martin Kraska  Am 13. Aug 2017 um 9:38 schrieb Gunter Königsmann gunter@...: >> Hmm, are you sure that filenames with UTF8 characters are displayed >> corectly in general? To judge by the comments here: >> https://superuser.com/questions/157207/onwindows7dirortreecantshowunicodecharactersevenstartingcmdwith >> it is a little bit tricky. A web search for "win7 utf8 characters in >> file names" seems to turn up some problems. >> >> I think draw2d will create a file named maxoutnnnn.gnuplot where nnnn >> is a process ID or something like that. If you look at that with a >> text editor, do you see the file name correctly? If the file name >> seems incorrect, can you verify that the bytes are correct somehow? > > The file will be in the directory indicated by maxima when typing > > maxima_tempdir; > > Another test would be: > > wxdraw2d(key="▶±ħ↝üØλΕφ",explicit(sin(x),x,1,10)); > > If the key is displayed correctly the file maxima passes to gnuplot > definitively contains unicode characters (no single codepage contains > greek, german and mathematical symbols). > ...in this case we can hope that gnuplot is the problem: Normally the > gnuplot programmers are quite fast and willing to help. > > Kind regards, > > Gunter. 