Evfuncs

- Add `evfuncs.segment_notes` function and tests for it
- Add Matlab script `make_oracle_data_for_tests.m` that generates "oracle" data from
actual functions written by Evren and saves in .mat files for comparison with output
from evfuncs

- Initial version after excising from hvc (https://github.com/NickleDave/hybrid-vocal-classifier/commits/0c50144d75e3a3205db82add8b48302edbbed511/hvc/evfuncs.py)
- Change `evfuncs.readrecf` to consistently parse values the same way, e.g. sampling frequency always returned as an int
- Convert tests to Python unittest format (instead of using PyTest library)
- Change defaults for `evfuncs.smooth_data` so it acts like `SmoothData.m` (and passes those defaults to `evfuncs.bandpass_filtfilt`)
- Write README.md with usage