|
CoreNEURON
|
Classes | |
| class | Instrumentor |
| Instrumentation infrastructure for benchmarking and profiling. More... | |
| struct | NullInstrumentor |
Typedefs | |
| using | InstrumentorImpl = detail::Instrumentor< detail::NullInstrumentor > |
Functions | |
| template<typename F > | |
| void | nrn_buildjacobian_thread (NewtonSpace *ns, int n, int *index, F const &func, double *value, double **jacobian, _threadargsproto_) |
| Calculate the Jacobian matrix using finite central differences. More... | |
| using coreneuron::detail::InstrumentorImpl = typedef detail::Instrumentor< detail::NullInstrumentor> |
Definition at line 285 of file profiler_interface.h.
| void coreneuron::detail::nrn_buildjacobian_thread | ( | NewtonSpace * | ns, |
| int | n, | ||
| int * | index, | ||
| F const & | func, | ||
| double * | value, | ||
| double ** | jacobian, | ||
| _threadargsproto_ | |||
| ) |
Calculate the Jacobian matrix using finite central differences.
Creates the Jacobian matrix by computing partial derivatives by finite central differences. If the column variable is nonzero, an increment of 2% of the variable is used. STEP is the minimum increment allowed; it is currently set to 1.0E-6.
| n | number of variables | |
| x | pointer to array of addresses of the solution vector elements | |
| p | array of parameter values | |
| func | callable that computes the deviation from zero of each equation in the model | |
| value | pointer to array of addresses of function values | |
| [out] | jacobian | computed jacobian matrix |
Definition at line 45 of file newton_thread.hpp.