Pyswi

==========================

- SWI calculation in swi_error_prop() decoupled form input uncertainty data availability
- quality flags normalized within swi_error_prop() and returned as percentages
- restarting calculation with stored parameters (gain_in/out)
- tests for above changes and swi_error_prop()/calc_swi_ts() result equivalence

==========================

- Includes an uncertainty characterization scheme from `Pasik et al. (2023, in review) <https://doi.org/10.5194/egusphere-2023-47>`_.
- Public release on pypi (``pip install pyswi``)
- Inclusion of iterative storage module

================================================

- Merged calc.py and iterative_swi.py into a single file
- Weights now optional in the original code (weighted functions removed)
- Unified variable names
- Updated the environment for py3, dropped py2 support

================================================

- Python compatibility >= 3.6
- Update pyscaffold v3.2.3

================================================

- Fixing performance bug in swi_ts by avoiding pandas.to_datetime function
- New simplified interface for swi_ts

================================================

- Adds processing_start/end to empty_data