What's new in PSICS 1.1.1
Mar 7, 2013
- On some platforms, access to the runtime arguments (the name of the xml file to run) was giving erroneous strings padded with junk rather than blanks. Trimming these strings therefore gave spurious names for the files in which the results were saved.
- An initialization problem in the core calculations for empty synapse populations caused an array overwrite and correspondfing nasty problems. It was assuming synapses[n_synapses] would always exist, wherease it doesn't if n_synapses=0 (array indexes running from 1 to n_synapses).
New in PSICS 1.09 (Mar 7, 2013)
- Increased the workspace for random numbers to support larger populations of stochastic channels. There should now be better error reporting too if the same problem occurs in future.
New in PSICS 1.08 (Mar 7, 2013)
- You can now have more than one DensityAdjustment element, so, for example, different channel types can be used on different regions of the cell. The DensityAdjustment elements are applied in sequence and each only affects the region in which both channel types are present.
New in PSICS 1.07 (Mar 7, 2013)
- Support for explicit sequences of synaptic events. Previously the ability to read an event sequence from an external file (EventSequence) was documented but not implemented. Now it is read and processed in Java and supported by the fortran core.
New in PSICS 1.05 (Mar 7, 2013)
- Removed unnecessary warning about not finding synapse populations: it wasn't finding them, but it also didn't need to as they were dereferenced independently after parsing the xml.