Channel combination functions currently assume that channels is the first dimension in passed in data, which is inconvenient and means that data must often be reorganised before calling the functions. It would be better if this was brought inside the function and that the dimension containing channels can just be passed in as an axis. It would also make sense for this to default to a value of -2, which is the default axis for raw data loaded by suspect.io functions.
Channel combination functions currently assume that channels is the first dimension in passed in data, which is inconvenient and means that data must often be reorganised before calling the functions. It would be better if this was brought inside the function and that the dimension containing channels can just be passed in as an axis. It would also make sense for this to default to a value of -2, which is the default axis for raw data loaded by suspect.io functions.