# E-fishBook

Database of data recorded in the group.



## Assumptions & Caveats:
* there is only a single subject in each dataset
* each recording contains a single cell
* RePro links to Cell and not to Dataset
* Question: should we store the sampling rate and hardware info in the database?

## FIXMEs
* repro must have foreign keys to dataset **and** subject
* Dataset finding samplerate in stimuli.dat can be improved
* Reading fi data from old-style directories
* maybe delegate this to Stimulus?
* implement alternative dataSources, somehow 
* stimulus contrast is not read from nix file (e.g. filestimulus) maybe a problem with all mutables?
* Add relacs_nix_version to dataset, could be used to decide how to read stuff, in particular, we will switch to DataFrames for most features after nixio version 1.5