search

mhoffman / kmos

kMC on steroids: A vigorous attempt to make lattice kinetic Monte Carlo modelling easier
http://mhoffman.github.com/kmos/
GNU General Public License v3.0
53 stars 34 forks source link

local_smart does not allow for conditions with an OR in the species field #38

Open jmlorenzi opened 8 years ago

jmlorenzi commented 8 years ago

In lat_int you can say

Condition(cood=some_coord, species='CO or NO')

but not in local_smart.

I can try to implement that if you think it is useful.

mhoffman commented 8 years ago

Hi, yeah

I am actually not sure it is so useful. It sure is sweet syntactic sugar but I think reintroducing it back into local_smart is quite intricate because local_smart is quite intricate. I think in many cases it would be easier to generate two processes for each case or use a Bystander in combination with your OTF backend. I think that is a more general approach and quite powerful, too! Right now I see more use-cases if we really iron-out your OTF backend and generate conditions like the one you mention on they way. Do you have a different viewpoint on this?

Best, Max.

jmlorenzi commented 8 years ago

I do think is more interesting to go on an improve otf. However, there are models in which local_smart is faster than either local_smart or otf (at least in it's current state). I suspect that allowing checks of several possible species right in case(spec1,spec2,...) in the search tree will be faster than configuring (potentally very many) extra processes... I also expected it not to be so difficult to put in. But maybe the risk of breaking something outweighs this relatively small advantage...

mhoffman commented 8 years ago

Ok, for you concern of breaking the backend I should put in some runtime tests for model using the existing syntax without OR. Give me a couple of days.