v0.0.4

SaI Krylov for matrix exponential, draft version
fix release v0.0.3

v0.0.2

• Magnus integrator can now be used to return error and error estimates for multiple time-steps. In one of the examples this is used to illustrate the asymptotic order of the different Magnus integrators and to verify error estimates.
• Subroutine to approximate exp(itA)v is now set in the propagator class.
• Krylov propagator can now be used to return error and error estimates for multiple time-steps or for all m and one time-step
• Krylov propagator can be used with fixed time steps to test performance, in this case, no adaptive settings are used

initial release

Initial release to pypi.