Alois Steindl

Hello, it seems that your matcont system is in some strange state. Try to call matcont clean or matcont clean reset as proposed right at the start of matcont.m Good luck Alois

Hello, your question is quite unspecific. I have no idea what you are trying to do. What happens at the Hopf bifurcation, do you have several pairs of imaginary eigenvalues? In that case it might be necessary to treat an imperfect system, where the Hopf bifurcation points are separated (one critical pair at a HBP). But that is just a wild guess based on your question.

Hello, MatCont isn't intended to compute this type of solutions, as far as I know. It wants to see generic types of bifurcations. You might try to reduce your system to the common level surfaces of H and F and use some BVP solver to calculate periodic solutions. Good luck Alois

Thank you very much for your quick and helpful response! The "set terminal" line was created by "save gnuplotfilename", after removing the "nopreamble", it works fine. The nopreamble entry is obtained by "show term". With kind wishes Alois

Thank you very much for your quick and helpful response! The "set terminal" line was created by "save <gnuplotfile>", after removing the "nopreamble", it works fine. The nopreamble entry is obtained by "show term". With kind wishes Alois</gnuplotfile>

Hello, the error messages in MatCont are sometimes hard to understand. I would guess, that at a certain point it couldn't find a solution. Maybe you should inspect the last curve and solution point and look, whether something looks strange. It might just be, that it encountered a higher degeneracy and failed to proceed. Sometimes it happened to me, that MatContm created an extremely large "test vector" and that caused strange intermediate results. So one could play with different settings in the...

Just a correction to my posts: It should read v1=T*vn and out(bp+2)= (w1*(fxhess'+2*S*sysjacv1))/(2*res); And of course: The excitation amplitude is called uhat (instead of gamma, as in my previous post).

Just a correction to my post: It should read v1=T*vn and out(bp+2)= (w1*(fxhess'+2*S*sysjacv1))/(2*res);