Ticket #55 (new enhancement)
Opened 3 years ago
What about an Interval description instead of a t_start/t_stop ?
| Reported by: | pierre | Owned by: | Pierre |
|---|---|---|---|
| Priority: | major | Milestone: | 0.2 |
| Component: | ALL | Version: | trunk |
| Keywords: | Cc: |
Description
You may have noticed, but an Interval class appears in the signals package. This is not used yet, but it will be very easy to implement it and to have a more powerful representation of the spike train and of all the related function. A SpikeTrain could be defined on an Interval (by default, if nothing is done, it will of course be [t_start, t_stop] and will not depend on the Interval dependency). But it will let you the possibility to have a real interval [0, 10], [20, 40], etc... Therefore, all the function like mean rate, activity movie, etc....should be able to be called on an Interval, and it could be very useful for example when you have several stimuli in a single trials, instead of doing a lot of time slices(). So what about an interval parameter instead of t_start, t_stop ? It would induce a change in the syntax, but not a big one. For Example SpikeTrain(times, t_start, t_stop) would be turned into SpikeTrain(times, interval). If of course interval = (t_start, t_stop), it will be compatible. So who will be interested by such a generalization ?
