Continuation simulations

[GoogleCodeExporter]

Describe the requested feature?
I think that when continue=>1 is set, t_start should default to the current 
time (as it is) and t_end should get set equal to t_start + t_end.  For 
example, in the following set of actions
#
simulate_ode(t_end=>100)
setConcentration("L()","L0")
simulate_ode(continue=>1,t_end=>10)
#
The 1st simulation should run from 0 to 100 and the 2nd simulation from 100 to 
110.  Currently, this code throws an error since t_start=100 > t_end=10 in the 
2nd simulate command.

This should be easy to do although care must be taken to handle cases where 
'continue', 't_start' and/or 't_end' are all defined.

How will the feature be useful to the system biology community?
Besides being more intuitive behavior, this will make simulations involving 
regularly-spaced actions (such as cell division) much easier to encode.

Original issue reported on code.google.com by lh64@cornell.edu on 7 Jun 2012 at 4:18

[GoogleCodeExporter]

I like the idea of supporting a 't_dur' argument for defining the simulation 
duration rather than changing the interpretation of t_end.  We just have to 
decide what to do if t_start, t_end and t_dur are all defined and t_end-t_start 
!= t_dur.  Probably easiest to give t_start and t_end precedence over t_dur.

Original comment by lh64@cornell.edu on 8 Jun 2012 at 9:29

[lh64]

Also, 'prefix' should be automatically set to the value from the previous simulation run. 'suffix' is also incompatible with 'continue', so if the suffix is changed BNG should throw an informative error.