A deep learning library for spiking neural networks which is based on PyTorch, focuses on fast training and supports inference on neuromorphic hardware.
Currently, when converting a list of events to a raster, using ChipFactory.events_to_raster, the timestamps will be aligned automatically such that the time of the first event is 0. This can be useful when the timestamps start with a high value and we only care about the timings after the first event.
However, often the timestamper is reset before recording events and hence the time of the first event can be of interested. I suggest providing the option of not aligning timestamps, in order to preserve this information.
Currently, when converting a list of events to a raster, using ChipFactory.events_to_raster, the timestamps will be aligned automatically such that the time of the first event is 0. This can be useful when the timestamps start with a high value and we only care about the timings after the first event.
However, often the timestamper is reset before recording events and hence the time of the first event can be of interested. I suggest providing the option of not aligning timestamps, in order to preserve this information.