Rapyuta

- Imported error ellipse support from a marvelous Python library of F. Galliano
+ `arrays`: `arrayize`, `ramp`
+ `plots`: `plotrange`, `Btau`, `Ctau`, `ellipse` and `SUE`
+ `plots.plotool`: `eplot`
- `plots.plotool`
+ Fixed the bug when any of `nrows` or `ncols` equals to 1, `self.axes` is 1D so `self.ax` cannot be attributed
+ Added `xisln` and `yisln` kw for `plot`
+ Added `close`
- `plots.pplot`
+ Fixed the bug when legend anchor kw appears tight layout is auto deactivated
- Added `tests/test_plots.py`
- Renamed `arrays.allist` to `arrays.listize`

- Rename package from `laputan` to `rapyuta` due to the French pun...thus it turns to the Ghibli origin!
- `plots.plotool`
+ Removed `set_font` and fixed font size and tick problems
+ Added `set_legend` which supports lengend boxes outside the axes
+ Renamed `set_border` as `set_fig` and added suptitle
+ `set_ax`
* Added font size kw
* Renamed `xlab` as `xlabel` (idem. for `ylab`, thus coherent to matplotlib)
* Changed `nonposx` default value from 'sym' to 'clip'
+ Note that all kw in `set_fig` and `set_ax` can also be executed individually with `self.ax.set_kw`

- Adopted Python `warnings.warn` for warnings in all libraries
- `arrays.closest`: Replaced error by warning raised and ignored kw `side` when it is not satisfied
- `imaging.concatenante` (Idem. for `respect.concat`): Removed wrange closest left/right side

- `imaging.imontage`: Corrected bugs in `coadd` Monte-Carlo uncertainty propagation

- `calib.intercalib`
+ Corrected `correct_spec` non-scalar gain/offset input
- `imaging.concatenante` (Idem. for `respect.concat`)
+ Added kw `keepfrag`, `cropedge`
+ Corrected indices with sided `closest`

- `arrays.closest`
+ Added kw `side` (default: None)
- `calib.intercalib`
+ Replaced `specorrect` kw `factor` by `gain`
+ Non-scalar `gain` and `offset`
+ Renamed `specorrect` as `correct_spec`
+ Extended `wcenter` to `read_filter` considering discrepancy between broad band offset and spectral offset
- Removed `maths.f_lin_p`