Setting Options for the Dynamics SolversΒΆ

Occasionally it is necessary to change the built in parameters of the dynamics solvers used by for example the qutip.mesolve and qutip.mcsolve functions. The options for all dynamics solvers may be changed by using the Options class qutip.solver.Options.

In [1]: options = Options()

the properties and default values of this class can be view via the print function:

In [2]: print(options)
Options:
-----------
atol:              1e-08
rtol:              1e-06
method:            adams
order:             12
nsteps:            1000
first_step:        0
min_step:          0
max_step:          0
tidy:              True
num_cpus:          0
norm_tol:          0.001
norm_steps:        5
rhs_filename:      None
rhs_reuse:         False
seeds:             0
rhs_with_state:    False
average_expect:    True
average_states:    False
ntraj:             500
store_states:      False
store_final_state: False

These properties are detailed in the following table. Assuming options = Options():

Property Default setting Description
options.atol 1e-8 Absolute tolerance
options.rtol 1e-6 Relative tolerance
options.method ‘adams’ Solver method. Can be ‘adams’ (non-stiff) or ‘bdf’ (stiff)
options.order 12 Order of solver. Must be <=12 for ‘adams’ and <=5 for ‘bdf’
options.nsteps 1000 Max. number of steps to take for each interval
options.first_step 0 Size of initial step. 0 = determined automatically by solver.
options.min_step 0 Minimum step size. 0 = determined automatically by solver.
options.max_step 0 Maximum step size. 0 = determined automatically by solver.
options.tidy True Whether to run tidyup function on time-independent Hamiltonian.
options.num_cpus installed num of processors Integer number of cpu’s used by mcsolve.
options.rhs_filename None RHS filename when using compiled time-dependent Hamiltonians.
options.rhs_reuse False Reuse compiled RHS function. Useful for repeatative tasks.
options.gui True (if GUI) Use the mcsolve progessbar. Defaults to False on Windows.
options.mc_avg True Average over trajectories for expectation values from mcsolve.

As an example, let us consider changing the number of processors used, turn the GUI off, and strengthen the absolute tolerance. There are two equivalent ways to do this using the Options class. First way,

or one can use an inline method,

Note that the order in which you input the options does not matter. Using either method, the resulting options variable is now:

In [3]: print(options)
Options:
-----------
atol:              1e-08
rtol:              1e-06
method:            adams
order:             12
nsteps:            1000
first_step:        0
min_step:          0
max_step:          0
tidy:              True
num_cpus:          0
norm_tol:          0.001
norm_steps:        5
rhs_filename:      None
rhs_reuse:         False
seeds:             0
rhs_with_state:    False
average_expect:    True
average_states:    False
ntraj:             500
store_states:      False
store_final_state: False

To use these new settings we can use the keyword argument options in either the func:qutip.mesolve and qutip.mcsolve function. We can modify the last example as:

>>> mesolve(H0, psi0, tlist, c_op_list, [sigmaz()], options=options)
>>> mesolve(hamiltonian_t, psi0, tlist, c_op_list, [sigmaz()], H_args, options=options)

or:

>>> mcsolve(H0, psi0, tlist, ntraj,c_op_list, [sigmaz()], options=options)
>>> mcsolve(hamiltonian_t, psi0, tlist, ntraj, c_op_list, [sigmaz()], H_args, options=options)