Version 2.2.7

This version adds some new features including:

• a method to Initialize a UVCal object from the metadata in a UVData object
• an option to reorder_blts to sort autos first

It also fixes a series of small issues, full changes detailed below and in our changelog:

What's Changed

• Adding a check to verify autos are real-only by @kartographer in #1110
• Remove distutils calls that are throwing deprecation warnings by @bhazelton in #1128
• Add option to reorder_blts to sort autos first by @bhazelton in #1127
• Make handling of types in Miriad extra_keywords more general by @plaplant in #1134
• Use time tolerances for matching times in UVFlag.to_baseline() by @jsdillon in #1133
• pin the RTD theme version to restore bullets in docs by @bhazelton in #1135
• Improve the documentation on redundancy finding, make keywords consistent by @bhazelton in #1132
• Use pytest < 7 because of conflict with pytest-cases by @bhazelton in #1138
• Increase version requirement for pytest-cases, unpin pytest by @bhazelton in #1140
• add a method to Initialize a UVCal object from the metadata in a UVData object by @bhazelton in #1035
• Tutorial and Docs restructuring by @bhazelton in #1139
• Make release 2.2.7 by @plaplant in #1141

Full Changelog: v2.2.6...v2.2.7

Version 2.2.6

This version adds some new features, including:

• Support for flexible spectral windows on UVCal objects.
• Support for wide-band calibration in UVCal
• Support for converting UVCal objects to use the future array shapes to allow users to upgrade their code before the shapes change.
• Support for flexible polarization setups on UVData objects, allowing for individual spectral windows to carry different polarization data.

This version fixes several bugs, the highest profile ones are:

• A bug where UVData.compress_by_redundancy sometimes produced incorrectly conjugated visibilities.
• A bug where lsts and lst_range got ignored when doing partial i/o with multiple files.

Full changes detailed below and in our changelog:

What's Changed

• Bring over kwargs from UVH5.write_uvh5_part() to UVData.write_uvh5_part() by @jsdillon in #1121
• Fix issue where lsts and lst_range was not properly being passed through uvdata.read() for multi-file reads by @jsdillon in #1122
• Fixing conjugation issue in compress_by_redundancy by @kartographer in #1123
• Add future_array_shapes and flexible spectral windows support to UVCal by @bhazelton in #1034
• Improved handling of MIR Data + flexible polarization for UVData by @kartographer in #1117
• update the changelog for v2.2.6 by @bhazelton in #1126

Full Changelog: v2.2.5...v2.2.6

Version 2.2.5

This version adds some new features, including:

• Dramatic improvement in memory performance in writing measurement set files and reading MWA correlator files.
• Support for reading the new MWAX correlator files.
• Defining scan numbers in Measurement Sets.
• A filename attribute to UVCal, UVBeam and UVFlag objects, matching the implementation in UVData objects.

This version fixes a number of bugs, the highest profile ones are:

• A bug that could cause incorrect data labelling when combining UVData objects with overlapping time-baselines/polarization/frequency channels that were ordered differently.
• Several measurement set bugs that caused problems when the files were used in CASA.

Dependency changes:

• h5py>=3.0
• astroquery>=0.4.4 (this is an optional dependency)

Full changes detailed below and in our changelog:

What's Changed

• Fix a bug in efield_to_power with only one feed by @bhazelton in #1075
• Update CI for external tests to get them working by @bhazelton in #1078
• added warning to uvcalibrate under x_orientation mismatch by @nkern in #1006
• mwa_corr_fits cleanup by @PyxieLouStar in #1079
• Fix expected shapes for data-like arrays on metadata only objects by @bhazelton in #1082
• fix doctests by requiring recent astroquery, updating tutorial output by @bhazelton in #1091
• Uvfits frame by @mkolopanis in #1086
• Handle None values in extra_keywords in UVH5 files by @plaplant in #1088
• Add check for UVH5 metadata only objects by @plaplant in #1092
• Add filename parameter to UVCal, UVBeam and UVFlag by @bhazelton in #1084
• Fix h5py deprecation warnings by @plaplant in #1097
• Write scan number for MS files by @e-koch in #1087
• Astroquery/JPL Horizons deprecation fix by @kartographer in #1099
• Suggesting write_ms when the number of ants > 256 by @kartographer in #1100
• Fixing an issue with UVData objects being incorrectly added by @kartographer in #1105
• MS-related bug fixes by @kartographer in #1101
• Mwax corr fits reader by @PyxieLouStar in #1090
• Testing if yaml change allows github tests to pass by @kartographer in #1113
• Fixes and performance improvements to the measurement set writer by @kartographer in #1112
• Add support for baseline coordinate suffixes in uvfits files by @bhazelton in #1106
• Add a CI to check for warnings, fix some tests by @bhazelton in #1118
• Measurement set TIME_CENTROID field fix by @kartographer in #1119
• Fix a bug in write_calfits when time_range is None by @bhazelton in #1116
• update the changelog for v2.2.5 by @bhazelton in #1120

Full Changelog: v2.2.4...v2.2.5

Version 2.2.4

This version fixes a bug in UVBeam.efield_to_power that was causing the Conda package builds to fail.

Full changes detailed below:

What's Changed

• Raise clear errors when trying to write metadata only objects to files by @bhazelton in #1072
• Fix conversion of non-crosspol power beams to real by @bhazelton in #1074

Full Changelog: v2.2.3...v2.2.4

Version 2.2.3

This version fixes a few minor bugs that were breaking Conda builds and adds OVRO-LWA to our known telescopes.

Automatically generated details below:

What's Changed

• Fixing automatically generated catalog IDs by @kartographer in #1067
• Add OVRO-LWA to known telescopes by @rlbyrne in #1060
• Increase the tolerance on uvbeam real_if_close call by @bhazelton in #1066
• Putting in a workaround for astroquery/JPL-Horizons related error by @kartographer in #1069

Full Changelog: v2.2.2...v2.2.3

Version 2.2.2

This version add a Measurement Set write method to UVData objects (UVData.write_ms) and methods to easily set data, nsamples and flags for baselines (set_data, set_flags, and set_nsamples) paralleling the popular get methods for these arrays. It also adds a new from_file class method for UVData and UVBeam objects to support one-line file reading into these objects.

Other improvements and bug fixes are detailed in the changelog.

Version 2.2.1

This version has an updated __eq__ method on the UVBase object which includes an allowed_failures keyword to specify parameters which are allowed to be unequal without causing the overall equality check to fail. For UVData, this keyword defaults to include the filename attribute.

Version 2.2.0

This release dramatically improves how phasing is done on UVData objects and adds support for multiple phase centers and non-sidereal phase centers. Reading and writing out data with multiple phase centers is supported for UVH5, UVFITS, MIR, and MIRIAD files. It also adds support for a filename attribute on UVData objects and support for getting and selecting data on UVData objects using LSTs.

Other improvements and bug fixes are detailed in the changelog.

Version 2.1.5

This release adds optional Van Vleck correction for MWA correlator files, a method for reordering frequencies on UVData objects and support for converting UVData objects to use the future array shapes to allow users to upgrade their code before the shapes change.

Other additions and bug fixes are detailed in the changelog.

Version 2.1.4

This release includes a major improvement in memory usage for reading MWA correlator FITS files and speed and memory improvements for finding and handling redundant baseline groups. See the changelog for full details of other bug fixes and improvements.