Many experiments contain some sort of experimental manipulation such as delivery of a stimulus that may be instantaneous or may last for a period of time. Despite this being essential to understand the data, there is currently no practical way to express this except to put a custom tag in every animal at every timepoint where something happened.
We should provide some guidance or perhaps an explicit recommendation on how to do this.
One option would be be to create a new top-level key, perhaps events, that would be an array of zero or more event entries that would have the structure ["arbitrary JSON key", 1.0] for an instant event or [{"id": 7, "arbitrary JSON":true}, 3.5, 4.5] for an event that takes some time to unfold. Subselection of time windows and merging of split files could then be handled in reference implementations for WCON.
Many experiments contain some sort of experimental manipulation such as delivery of a stimulus that may be instantaneous or may last for a period of time. Despite this being essential to understand the data, there is currently no practical way to express this except to put a custom tag in every animal at every timepoint where something happened.
We should provide some guidance or perhaps an explicit recommendation on how to do this.
One option would be be to create a new top-level key, perhaps
events, that would be an array of zero or more event entries that would have the structure["arbitrary JSON key", 1.0]for an instant event or[{"id": 7, "arbitrary JSON":true}, 3.5, 4.5]for an event that takes some time to unfold. Subselection of time windows and merging of split files could then be handled in reference implementations for WCON.