GitLab

- logged commandline can contain full path to exectable, so
  cmdline-comparison must be agnostic to that
- fixes failing test against pip-installed lalsuite

- fix test that can otherwise fail (observed on MacOSX)

- supply consistent flake8 settings in setup.cfg

- tests currently fail

Closes #12

 This removes the convergence testing ideas previously implemented
 (currently juts commented, but later to be fully removed). These are
clearly not useful without further study, which in itself would be a
better time to develop an implementation.

Fixes #4 .

The variable name was misleading and hence renamed

This renames the input sftfilepath to sftfilepattern and adds
documentation on how that should be used, i.e. a colon separated list of
wildstring or exact matches. In globbing for all matches, the colon
split is added in. sftfilepath is still used by `Writer` since
an exact path is known.

The command line argument is matched to formatting used by MFD to
reconstruct the argument (using .16g) formatting. Ideally, all REAL8
should use this formatting.

- restructures the method into several different functions.
- implements stopping if convergence is met a prespecified number of times
- adds documentation
- will plot the convgerence on the RH axis of the walkers plots

Convergence testing based on the Rubin-Gelman statistic is implemented
via `setup_convergence_test`. This does not truncate the search yet, but
simply adds a trace to the walkers plots for testing

Now forces use of dtglitch = None for no glitches

Transforms the single pyfstat.py into a python module splitting the
relevant code into separate sub-files in pyfstat. This should result in
improved readability.

1) Moves the actual estimation to external function since it is not
contingent on using the MCMC methods etc
2) Change notation to V, Vsky and Vpe
3) Fix some table outputs (Tcoh was forcing integer)
4) Rename of names -> detector_names
5) Adds tests
6) Writes a simpler table if the sky is not searched over

tstart tend refers to signal properties, so is only used when relevant.
Otherwise use the LAL parameter min/maxStartTime

It seems the tests have been failing, but only falling out of range by
~0.11 rather than 0.1. This increases the allowed range to 0.2