[go: up one dir, main page]

User Activity

  • Posted a comment on ticket #3481 on Maxima -- GPL CAS based on DOE-MACSYMA

    Hi Robert, thanks for your help. Sorry to be a nuisance. Some programming languages offer warnings when functions (particularly built-in ones) are accidentally redefined, would there be a way to turn this on in Maxima?

  • Posted a comment on discussion ngspice-devel on ngspice

    SuiteSparse won't directly compile for windows. But https://github.com/jlblancoc/suitesparse-metis-for-windows builds fine on windows and for CUDA. As far as I am aware, cuSPARSE is a little behind cuda-enabled SuiteSparse in performance and features. How trivial would it be to use the relevant elements of SuiteSparse, built for windows with cuda, instead of cuSPARSE?

  • Created ticket #3481 on Maxima -- GPL CAS based on DOE-MACSYMA

    ~ used with improper arguments: [-3,4,-1] [1,-3,2]

  • Posted a comment on discussion ngspice-devel on ngspice

    Having a bit of trouble here. According to the manual, the syntax is correct. But it appears that + signs have already been parsed out of the netlist when translatepoly is called. So when translatepoly adds in plus signs, to form an expression for the polynomial, ngspice returns an error. Full details here: https://sourceforge.net/p/ngspice/discussion/127605/thread/5356888d/#f512/9a38

  • Modified a comment on discussion ngspice-devel on ngspice

    I have followed Holger's example in translatepoly and used cieq. This is included in an optional double check in interpretpoly, enabled by defining DOUBLE_CHECK_POLY TRUE. git diff --cached gives the changes below. diff --git a/src/frontend/translatepoly.c b/src/frontend/translatepoly.c index cf223594c..3de5bde0f 100644 --- a/src/frontend/translatepoly.c +++ b/src/frontend/translatepoly.c @@ -12,6 +12,7 @@ Author: 2018 Thomas P. Dye #include "translatepoly.h" #define TRANSLATEPOLY_REPLACE TRUE +#define...

  • Posted a comment on discussion ngspice-devel on ngspice

    I have followed Holger's example in translatepoly and used cieq. This is included in an optional double check in interpretpoly, enabled by defining DOUBLE_CHECK_POLY TRUE. git diff --cached gives the changes below. diff --git a/src/frontend/translatepoly.c b/src/frontend/translatepoly.c index cf223594c..3de5bde0f 100644 --- a/src/frontend/translatepoly.c +++ b/src/frontend/translatepoly.c @@ -12,6 +12,7 @@ Author: 2018 Thomas P. Dye #include "translatepoly.h" #define TRANSLATEPOLY_REPLACE TRUE +#define...

  • Posted a comment on discussion ngspice-devel on ngspice

    This fix is not good enough. It looks like "POLY" and "poly" would both make it through previous functions unaltered. I need to explore a case insensitive approach.

  • Posted a comment on discussion ngspice-devel on ngspice

    Changing to curly brackets does not remove the error: (gdb) run Starting program: /usr/local/bin/polyngspice -b polytest1.sp [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". Circuit: polytest1 Error on line 0 : be.x1e.e1 v = {{1.00000000000000000e+00} + {{1.00000000000000000e+00}*{v( nvctrl , 0 )}} + {{1.00000000000000000e+00}*{v( nvctrl , 0 )}*{v( nvctrl , 0 )}}} unknown parameter (+) Error on line 0 : g.x1g.g1 nout1g 0 POLY(...

View All

Personal Data

Username:
bikeviolin
Joined:
2017-07-05 18:01:05

Projects

  • No projects to display.

Personal Tools