Page 1 of 1

Multiple hoc files, single model

Posted: Thu Apr 12, 2012 2:54 pm
by Qroid_montreal
Beyond conceptual clarity, is there a motivation to split hoc code into multiple hoc files? For example, could all the hoc files found in the example here (http://www.neuron.yale.edu/phpBB/viewto ... f=28&t=168) be combined without affecting the implementation of the model?

Re: Multiple hoc files, single model

Posted: Thu Apr 12, 2012 9:35 pm
by ted
In principle the answer is "yes." There are other principles, however, including most pertinently:
1. all code is garbage until proven otherwise
2. the principal task in computational modeling is establishing conceptual control, that is, being sure that there is a close match between what you think is in the computer, and what is actually in the computer
3. the only thing worse than trying to read and understand your own old code, is trying to read and understand someone else's old code
4. you're free to do what you like and see if it works
5. if you break it, you debug it