Changelog

v2.0.4 
- Normalized python code (spaces).
- Added new features to VGOS mode (UVTaper, multi-scan fitting, etc.)
- Corrected bugs in PolGainSolve (bad rate estimates degraded solution quality).

v2.0.3
- Synchronized ALMA/non-ALMA buildables; python codes still not unified
- This version appears to reproduce the ALMA behaviors of v1.8.5

v2.0b
- Improved the GCPFF algorithm. Implemented the interface for CASA 6.

v1.9.0
- Added new standalone interface (i.e., free of CASA!!!).

v1.8.5
- interim successor to be tagged in DiFX 2.7(.1)
- includes the getAmpRatio fix in _PolConvert
- updates CalTable.* git version to DiFX.
- a number of refactoring changes not yet included

v1.8.4
- finishing Py3 conversion with updates to make/configure within DiFX
- includes a start on tools to slice and dice sample datasets

v1.8.3
- interim successor to v1.8.1 for the DiFX branch of PolConvert
- (1.8.2 was developed on launchpad present with the git repo creation)

v1.8.2
- Improved GFF for rate estimates (now, it is done on all corr. products).
- Parallactic angle can now be corrected in converted antennas.
- Adapted the pulse-cal code for multi-file correlations.
- Cleaned up the code.
- Fixed problem in ChiSq code when the scans weren't in time order.
- Added sample EU-VGOS scripts.

v1.8.1
- improved support for FITS-IDI
- fixed some buggy parallactic angle code relevant to single dish

v1.8.0
- first cut at portage to Python 3

v1.7.9
- added checkpolconvert.py
- prepolconvert.py and drivepolconvert.py execute in py3 environment
- added comparepolconvert.py
- do not generate the FRINGE binary files for IFs that are not plotted

v1.7.8

- fixes for calibration table flexibility (getcol replaced by getcell)

v1.7.6

- New options to estimate cross-gains from station metadata (pcals).
- New configuration keywords to fine-tune non-ALMA cases (VGOS).
- Synchronized with DiFX 2.5.3 PP scripting for 2018
- Updated build to respect numpy includes

v1.7.5

- set zero weight on unphased APP subscans (phases not updated)
- continued non-ALMA development

v1.7.4

- Added some protections for problematic antennas in ALMA data tables
- Support for polarized calibrators (for non-ALMA observations).

v1.7.3

- minor bug fixes

v1.5 - v1.7

- antab for ALMA is now correct; missing cross-pol autocors for ALMA now fixed
- general expansion of capabilities in response to EHT issues and
- partial support for non-ALMA linear pol sites (EVN, &c.)
- autocrosscorrs now fixed (for GMVA)

v1.4

- Bandpass mode for the cross-polarization gains.
- Parallactic-angle computation (to enable the use of RR and LL in the X-Y gain estimate).
- Global Cross-Polarization Fringe Fitting, to find X-Y gains.
- New C++ module to solve for gains.

v1.3

- Returns cross-polarization gains in a list (to apply them to XYadd and XYratio).
- Estimate (and application) of multi-band delay

v1.2-r7

- Gain files in ANTAB format.
- Proper handling (by interpolation) of flagged gains.

v1.2-r8

- changes for cleaner compilation
- adjustments to task_polconvert.py for shape variations in cal tables

v1.2-r9

- Corrected bug in CalTable.
- Automatic selection of spw (if spw is set to <0 ).

v1.2-r10

- Several IFs can be plotted at once.
- Added plot of amplitude vs. IF number.
- Added "self" interpolation (useful for amplitude solutions).
- Plots and diagnostic files are now saved in directories.

v1.2-r11
- Added XYratio
- Added ability to drop XY phase solution

v1.2-r12
- Plots are generated in off-line mode.

v1.2-r13
- Modified for CASA 4.7

TODO: 

- Fitting of XY-phase with a simple delay model (per subband).
- BETTER LEAST-SQUARES MINIMIZER FOR THE CROSS-POLARIZATION GAIN ESTIMATES.

Version numbers are captured (and should be sync'd in):
    TOP/Changelog
    Changelog
    polconvert.xml
    task_polconvert.py
    PP/runpolconvert.py
    (and the single dish files...)