assumes the csv files have a certain structure in the columns: [metadata for experiment setup][channel for response recording][bins of time series responses], documented in README, but you don't need to know the names/numbers of metadata items or times, only the name for the channel column (GRN in our case)
made unit tests run on branches other than main
created dataset class to represent a whole csv file, contains multiple instances of timeseries objects to represent each row in the csv, and has global metadata for bookkeeping and to be sure each row is consistent.
added functions in timeseries and dataset classes to compare 2 objects, add/remove a timeseries from a dataset, and to rebin a timeseries or all the timeseries in a dataset.
all these things also added to ReadtheDocs
updated the read_data script to produce a dataset object rather than a list of timeseries objects
updated run script command line arguments to include e.g. the time interval used by default in the csv (e.g. 100ms)
Updates to the data structure: