All notable changes to this project will be documented in this file.
- ATA has been added to the list of known telescopes.
- New keyword
check_kw
tonew_uvdata
that lets the user specify options to the checking of the created object. - New keyword
all_times_unique
tocalc_app_coords
that short-cuts the calculation when it is known that all input times are unique. - New convenience methods on
BeamInterface
to simplify the handling of analytic vs UVBeam objects. - Added support for partial read for MWA correlator FITS files.
- Added
antenna_names
,time_range
,lsts
andlst_range
parameters toUVFlag.select
to match UVData and UVCal select methods.
- Made it possible to not return the
interp_basis_vector
array from beam interpolators.
- Initialization of analytic beams with single feed.
- New test failures related to incorrect tolerances for floating point comparisons in tests. Floating point comparisons have been updated throughout the code base.
- New analytic beam classes: AiryBeam, GaussianBeam, ShortDipoleBeam, UniformBeam that support evaluating either the efield or power beam in any direction and frequency.
- A new BeamInterface class that provides a unified interface to UVBeam and analytic beam objects to get beam responses in any direction and frequency (via interpolation or evaluation as appropriate).
- New optional spatial interpolation method,
interpolation_function="az_za_map_coordinates"
that improves the linear interpolation speed for data inaz_za
coordinates. - New UVParameter
pol_convention
onUVData
andUVCal
. This specifies the convention assumed for converting linear to stokes polarizations -- either "sum" or "avg". Also added touvcalibrate
to apply from theUVCal
to theUVData
. - New
ignore_telescope_param_update_warnings_for
function that globally ignores warnings for specific telescopes. - New
.telescope
cached property on theFastUVH5Meta
object that auto-creates a telescope object from the metadata. UVData.get_enu_data_ants
method to get east, north, up positions only for antennas with data.
- Updated minimum optional dependency versions: lunarsky>=0.2.5, pytest>=8.2.0
- A bug in the MWA beam reader that resulted in the wrong polarization response (the azimuthal-aligned response was swapped with the zenith angle-aligned response).
- A bug in reading UVH5 files with antenna names saved as variable length strings that was introduced in v3.0.0.
- Several new methods on
Telescope
objects, including the classmethodsfrom_known_telescopes
andnew
to instantiate new objects and theget_enu_antpos
method to get antenna positions in East, North, Up coordinates. - Support for writing "MODEL_DATA" and "CORRECTED_DATA" columns has been added to the
UVData.write_ms
method. - Support for "flexible-Jones"
UVCal
objects -- where different spectral windows can contain different Jones vectors/polarization information -- has been added, along with methods for convertingUVCal
objects to and from flex-Jones format (remove_flex_jones
,_make_flex_jones
, andconvert_to_flex_jones
). Support for flex-Jones has been added toMSCal
andCalH5
subclasses. - A new class called
MSCal
has been added as a subclass ofUVCal
, which adds support for reading and writing of CASA Measurement Set gains, bandpass, and delays tables. To support this, several new optional parameters have been added to theUVCal
class, includingphase_center_catalog
,phase_center_id_array
,antenna_diameters
,scan_id_array
, andref_antenna_array
(which affords the ability to record a different reference antenna per time-axis entry). - A new HDF5 file format for calibration solutions called
calh5
, which supports writing out all types of UVCal objects and supports partial reads (select on read). - A
UVCal.get_lst_array
method similar to theget_time_array
method that either returns the LST of the mean of the start and stop time for each time range or the lst_array (if there's a lst_array and no lst_range). - A
UVCal.get_time_array
method that either returns the mean of the start and stop time for each time range or the time_array (if there's a time_array and no time_range). - Added new keyword handling for v.6 of the MIR data format within
MirParser
.
- Updated minimum dependency versions: scipy>=1.8, setuptools_scm>=8.1
- Updated minimum optional dependency versions: lunarsky>=0.2.4
- Restructured
utils.py
into a new submoduleutils
with functions split among several submodules. Utility functions used widely are still available from pyuvdata.utils, although this access pattern is deprecated for some of them. - Modified
UVBeam.interp
to speed up processing whencheck_azza_domain=True
. - Updated minimum dependencies: setuptools>=64, setuptools_scm>=8.0
- Restructured to a
src
layout. This should not affect most users, but thecheck_warnings
function has moved frompyuvdata.tests
topyuvdata.testing
. - Most packaging information has now moved to the
pyproject.toml
. - Future array shapes are now the only supported standard for
UVData
,UVCal
,UVFlag
,UVBeam
classes. UVCal
objects where the cal type is delay must now havewide_band=True
.- The
flex_spw_id_array
is now required forUVCal
objects werewide_band=False
. UVCal
objects withwide_band=True
may only have thefreq_range
parameter set and thefreq_array
andchannel_width
parameters. Ifwide_band=False
, then thefreq_array
andchannel_width
parameters must be set, while thefreq_range
parameter must be left unset.- For
UVCal
objects, only one of the time-based parameters (time_array
andtime_range
) may be set on a given object. - Telescope-related metadata (including antenna metadata) on
UVData
,UVCal
andUVFlag
have been refactored into aTelescope
object (attached to these objects as thetelescope
attribute) and most of the code related to these attributes from the three objects has been consolidated on theTelescope
object. These attributes are still accessible via their old names on the objects, although accessing them that way is now deprecated. Note that the telescope locations are now stored under the hood as astropy EarthLocation objects (e.g.UVData.telescope.location
) (or as MoonLocation objects for simulated arrays on the moon iflunarsky
is installed). - The
UVData.new
andUVCal.new
methods now accept a telescope object for setting the telescope-related metadata (the older parameters are still accepted but deprecated.) - When FHD calibration solutions are read in, the
time_range
is now set to be one quarter of an integration time before and after the earliest and latest times respectively. This is a change from extending it by half an integration time to prevent apparent overlaps in neighboring calibration solutions due to jd precision loss. - Improved performance for
antpair2ind
and_key2inds
by using theblts_are_rectangular
parameter, and also by caching the results. To improve performance of the cache, the resulting indices are returned as slices whenever possible. - Changed the way files are passed for reading FHD files. Rather than one list including
all the types of files, each file type (e.g. params, obs, layout) must be passed to
their own parameter. This makes the API for reading FHD data files more similar to the
API for reading FHD cal files. Also removed the
use_model
parameter because it is no longer needed (only data or model files should be passed as visibility files). - Require keyword arguments rather than allowing for passing arguments by position for functions and methods with many parameters.
- Only one of
time_array
andtime_range
(and similarlylst_array
andlst_range
) can be set on a UVCal object. - If
time_range
is set it must be 2D with a shape of (Ntimes, 2) where the first axis gives the number of different time solutions and the second axis gives the start/stop times for those solutions.
- The
future_array_shapes
attribute onUVBase
objects has been deprecated, as pyuvdata now exclusively uses future array shapes. - The
use_future_array_shapes
keyword in several different class methods, as well as theuse_future_array_shapes
method onUVBase
objects. - Accessing the telescope-related metadata through their old attribute names on
UVData
,UVCal
andUVFlag
rather than via their attributes on the attachedTelescope
object (e.g.UVData.telescope_name
->UVData.telescope.name
andUVData.antenna_positions
->UVData.telescope.antenna_positions
). - Passing telescope-related metadata as separate parameters to
UVData.new
andUVCal.new
rather thanTelescope
objects. - The
UVData.get_ENU_antpos
method in favor ofUVData.telescope.get_enu_antpos
. - The
Telescope.telescope_location
andTelescope.telescope_name
attributes in favor ofTelescope.location
andTelescope.name
. - The
get_telescope
function in favor of theknown_telescope_location
function and theTelescope.from_known_telescopes
classmethod. - The KNOWN_TELESCOPE dict in favor of the
known_telescope_location
function and theTelescope.from_known_telescopes
classmethod.
- Fixed a minor bug in
utils._check_freq_spacing
that raised issues which raised an error when evaluating a "flex-spw" dataset where Nspws was 1. - Fixed a bug in
UVBase
whereallowed_failures
was being ignored if a parameter hadrequired=True
set. - Fixed a bug where selection and addition/concat methods did not operate correctly on
flex-spw
UVCal
objects whereNspws>1
.
- Support for current array shapes has been removed on
UVData
,UVCal
,UVFlag
, andUVBeam
objects, as well as theuse_current_array_shapes
method for these classes. - The
flex_spw
attribute has been removed onUVData
andUVCal
objects. - Support for using the old phasing attributes (
phase_center_ra
,phase_center_dec
,phase_center_frame
,phase_center_epoch
,phase_type
, andobject_name
) inUVData
. - Support for handling of the
input_flag_array
parameter forUVCal
objects. - Support for providing capitalized variants of
UVBeam.feed_array
. - Support for 'unknown' cal type in
UVCal
objects.
- numpy c-api for 2.0 compatibility
- numpy.string_ calls to numpy.bytes_, np.in1d to np.isin and a few other changes for numpy 2.0 compatibility.
- Fixed #1445 in which in some corner cases the
antpairs
array was not correctly interpreted as int.
- updated minimum cython dependency to 3.0.
- Updated minimum optional dependency versions: astropy-healpix>=1.0.2
- Fixed a bug where UVFITS header parameters RDATE and DATE-OBS were being improperly set.
- Fixed a bug with calculating Miriad antenna and baseline numbers, which properly accounts for MIRIAD antennas being 1-indexed not 0-indexed.
- Fixed a bug writing UVData objects to UVFITS formats with extra phase center catalog entries resulted in an error on read.
- Fixed a bug where calling
UVData._set_app_coords_helper
with extra phase center catalog entries resulted in an error. - Bug in which
correct_cable_len
defaulted toNone
instead ofTrue
forread_mwa_corr_fits
.
- Support for new lunar ellipsoids in
lunarsky
version 0.2.2. - Added new keyword handling for v.6 of the MIR data format within
MirParser
.
- Updated minimum dependency versions: astropy>=6.0, h5py>=3.4, numpy>=1.23, pyerfa>=2.0.1.1, scipy>=1.7.3, python>=3.10
- Updated minimum optional dependency versions: hdf5plugin>=3.2.0, pytest>=6.2.5, python-casacore>=3.5.2, pyyaml>=5.4.1, lunarsky>=0.2.2
- Speed up for reading MWA correlator FITS files into
UVData
. - Now compatible with pytest>=8.0 but require pytest-cases>=3.8.3.
- Fixed a bug where SMA-based coordinates were recorded as FK5 for sidereal sources (should be ICRS).
- Bug in which passing 'ee' or 'nn' pols to UVData.new() would error, even if
x_orientation
was passed.
- Added a new redundancy finding algorithm based on the HERA gridding algorithm. This new algorithm is used by default but the older clustering algorithm is still available.
- Added support for initializing
MirParser
andMirMetaData
objects fromPath
objects. - Added a brief tutorial on how to convert SMA MIR data into MS format (with some notes on SMA specific keywords than can be used).
- Added the
Mir.generate_sma_antpos_dict
method for reading in SMA-formatted antenna position files into a dict (which can be passed toUVData.update_antenna_positions
). - Added handling for older formatted (so called "ASIC-era") MIR files from SMA, as well as more recent file formats that preceded pyuvdata support (circa 2020, v1/v2 of the MIR file format).
- Added the ability to save/load selection masks in
MirParser
objects. - Fixed a bug when trying to write a MS with a legacy array shape (but note legacy shape will be deprecated in upcoming releases).
- Improved support for MIRIAD UV files, and added
use_miriad_convention
flag to the UVFITS writer code, so files with the MIRIADBASELINE
convention can be created. - A new
freq_interp_kind
parameter toUVBeam.interp
,UVBeam._interp_az_za_rect_spline
andUVBeam._interp_healpix_bilinear
to allow the frequency interpolation specification to be passed into the methods. Note this defaults to "cubic" rather than "linear" (the old default for the attribute of the same name on UVBeam objects) because several groups have found that a linear interpolation leads to nasty artifacts in visibility simulations for EoR applications. - A new
UVBeam.new()
method (based on new functionnew_uvbeam
) that creates a new, self-consistentUVBeam
object from scratch from a set of flexible input parameters. - Added a the
UVData.update_antenna_positions
method to enable making antenna position updates with corresponding updates the uvw-coordinates and visibility phases. - Added a switch to
UVData.write_ms
calledflip_conj
, which allows a user to write out data with the baseline conjugation scheme flipped from the standardUVData
convention. - Added the
utils.determine_pol_order
method for determining polarization order based on a specified scheme ("AIPS" or "CASA"). - Added a
check_surface_based_positions
positions method for verifying antenna positions are near surface of whatever celestial body their positions are referenced to (either the Earth or Moon, currently).
- Made the determination of whether or not to create a flex-pol dataset when reading in a MIR more robust (particularly with pre-V3 data formats).
- Reading in of MIR data sets into
UVData
objects will now use pyuvdata-calculated values forlst_array
via (UVData.set_lsts_from_time_array
) instead of those read in from the file due to known precision issues in the later (~25 ms), so long as the two agree within this known precision limit. UVFlag.to_baseline
andUVFlag.to_antenna
are now more robust to differences in antenna metadata sorting.- Made
MirParser
more robust against metadata indexing errors. - Made
MirParser
handling of autocorrelation data more robust. - Improved hooks for handling and loading of COMPASS solutions into
MirParser
objects. - Fixed handling of weights to be consistent with tsys scaling method in
MirParser
. - Changed
MWACorrFits.corrcorrect_simps
method to use thescipy.integrate.simpson
method rather than thescipy.integrate.simps
method to fix deprecation warnings. - added support for python 3.12, dropped support for python 3.8.
- Updated minimum dependency versions: pyyaml>=5.3
- Changed
UVData.write_ms
to sort polarizations based on CASA-preferred ordering. - Added some functionality to the
utils._convert_to_slices
method to enable quick assessment of whether an indexing array can be replaced by a single slice. - Increased the tolerance to 75 mas (equivalent to 5 ms time error) for a warning about
values in
lst_array
not conforming to expectations forUVData
,UVCal
, andUVFlag
(was 1 mas) inside ofcheck
. Additionally, added a keyword tocheck
enable the tolerance value to be user-specified. - Changed the behavior of checking of telescope location to look at the combination of
antenna_positions
andtelescope_location
together forUVData
,UVCal
, andUVFlag
. Additionally, failing this check results in a warning (was an error).
- The
freq_interp_kind
attribute on UVBeams. - The
spw_array
andNspws
attributes on UVBeam objects. Also theunset_spw_params
andset_spw_params
parameters to theuse_future_array_shapes
anduse_current_array_shapes
methods on UVBeam objects. - Upper case feed names (e.g. "N" or "E") in UVBeam.feed_array. This was never fully tested and didn't work properly.
- Having
freq_range
defined on non-wide-band gain style UVCal objects. - Having
freq_array
andchannel_width
defined on wide-band UVCal objects.
- A bug where
time_array
was not being correctly calculated for older (pre-V3) MIR data formats. - A small bug in UVFlag that could occur when reading in an older UVFlag HDF5 file with missing antenna metadata.
- A small bug (mostly affecting continuous integration) that threw an error when the
IERS service was down and when tests were using
test.check_warnings
to look for no warnings thrown (i.e., wherematch=None
). - A couple of small bugs related to handling of the
freq_range
parameter in thereorder_freqs
and__add__
methods onUVCal
.
- Added a
uvw_track_generator
method withinutils
for calculating the expected uvws (and a few other values) without needing to instantiate a wholeUVData
object. - Added a convenience function called
compare_value
inUVParameter
that enables one to do value checking with tolerances accounted for. - New
mwa_metafits_file
andtelescope_name
optional parameters toUVFlag.read
andUVFlag.__init__
to help with setting telescope metadata for old UVFlag files that are missing it. - MWA antenna metadata to our known telescopes to allow them to be filled in for old
UVFlag files. This is a stopgap solution, a better approach is to pass an MWA metafits
file to the new
mwa_metafits_file
parameter. - Support for recarrays in
UVParameter
objects and inUVBase
, needed for pyradiosky. - Support for setting the astrometry library for various object methods including
set_lsts_from_time_array
, file read methods and others. - Properly round-trip the telescope frame through UVH5, UVFITS and MS files.
- A bug in apparent coordinate calculation that resulted in small errors/loss of
precision due to the way times were passed when using
erfa
astrometry library. - A small correction due to polar drift for LST calculation when using the
erfa
astrometry library. - Fixed a bug in
utils.calc_app_coords
that occurred when supplying an astropyTime
object for thetime_array
argument. - Fixed a bug in
UVData.write_ms
that caused the writer to crash if prior measurement set history was not formatted in the currently expected fashion.
- Support for multiple spectral windows in
UVData.frequency_average
, including a new parameterrespect_spws
for controlling whether averaging crosses spectral window boundaries. - Better handling in
UVData.frequency_average
when averaging by a number of channels that does not divide evenly into the number of channels in each spectral window. - Compatibility with Python 3.11
- The following
UVFlag
parameters are now required:Nspws
,channel_width
,spw_array
,telescope_name
,telescope_location
,antenna_names
,antenna_numbers
, andantenna_positions
. - The
quality_array
on UVCal objects is no longer required.
- The
sky_field
attribute onUVCal
.
- A bug in LST calculation that led to small discontinuities in LSTs calculated using
the
erfa
ornovas
astrometry libraries. - Error when setting
UVBeam.freq_interp_kind
to an integer. - Error when reading
mwa_corr_fits
files from the new MWAX correlator
- The
phase_uvw
andunphase_uvw
utility methods associated with old style phasing. - The
with_conjugates
option to theget_baseline_redundancies
utility method in favor of theinclude_conjugates
option. - Support for the
interpolation_function
attribute on UVBeams. - Support for the "unphased" catalog type in UVData.phase_center_catalog in favor of the "unprojected" catalog type.
- The
UVData.unphase_to_drift
method in favor ofUVData.unproject_phase
method. - Support for using old style phasing on UVData objects (the
UVData.fix_phase
method to fix datasets that were phased with the old style of phasing remains as well as thefix_old_proj
option on read for uvfits, miriad and uvh5 file types). - Phasing related keywords in the following
UVData
methods:__add__
,__iadd__
,fast_concat
,read
andfrom_file
. Multiple phase centers are now fully supported so datasets can always be combined without phasing. If desired, the full object can be rephased to a single phase center afterwards. - The
use_cotter_flags
andflag_small_sig_ants
options to theUVData.read
andUVData.from_file
methods for MWA correlator fits files in favor of theuse_aoflagger_flags
andflag_small_auto_ants
options respectively. - The
spoof_nonessential
option toUVData.write_uvfits
method as the previously spoofed parameters are now properly calculated.
- New dependency
docstring-parser
to enable dynamic docstring generation for methods that mostly wrap other methods. - A new
UVCal.new()
method (based on new functionnew_uvcal
) that creates a new, self-consistentUVCal
object from scratch from a set of flexible input parameters. - A new
UVData.new()
method (based on new functionnew_uvdata
) that creates a new, self-consistentUVData
object from scratch from a set of flexible input parameters. - A new
fast_concat
method onUVCal
. - A new generic
read
method onUVCal
that, like theread
methods on our other objects, supports all file types and a newfrom_file
class method to allow one-line reading of UVCal files.
- The
input_flag_array
attribute onUVCal
. - Support for the 'unknown' cal_type in UVCal.
- Reading in multiple files to
UVCal
using file-type specific read methods (e.g.read_calfits
) in favor of the genericread
method.
- Removed error when
time_axis_faster_than_bls=True
andNtimes=1
. In the case ofNtimes=1
, it is inconsequential whether time-axis moves first or not, so it should not be an error.
- The
catalog_name
keyword has been added to theUVData.read
andUVData.select
methods to allow users to select on source name.
- The keywords for
UVData.read
have been changed for MIR-specific values to more generally match those founds with other data types. - The
UVData.read_mir
method has been overhauled to mak reading in of MIR data more efficiency in memory/processor usage.
- Frequency frame information for MS datasets is now correctly recorded as "TOPO".
- A new
FastUVH5Meta
object that enables quick partial reading of uvh5 file metadata. This is used by default in theUVData.read_uvh5
method to initialize aUVData
object, without any apparent change for the user. To get the benefits of the new class, use it directly to interface with the metadata portion of a.uvh5
file.
- Updated numpy requirements to
>=1.20
. - Updated scipy requirements to
>=1.5
.
- Handling of antenna_names and antenna_numbers in
read_fhd
andread_fhd_cal
.
- A new
interpolation_function
parameter toUVBeam.interp
andUVBeam.to_healpix
to allow the function name to be passed into the methods, with sensible defaulting. - Support for selecting on phase center IDs, including on read.
- Several new attributes to UVFlag:
telescope_name
,telescope_location
,antenna_names
,antenna_numbers
,antenna_positions
,channel_width
,spw_array
,flex_spw_id_array
, enabling support for flexible spectral windows. All of these exceptflex_spw_id_array
will eventually be required for all UVFlag object types, as willNspws
(which used to only be required for baseline and antenna types). These parameters all already exist on UVData and UVCal objects and UVFlag objects initialized from them will inherit them. UVFlag objects will also inherit these parameters when they are converted between types using UVData and UVCalobjects. - The
UVFlag.set_telescope_params
method, similar to the ones on UVData and UVCal, to set several of these new parameters.
- Increases the h5py minimum version to 3.1.
- Increase the minimum compatible version for lunarsky to 0.2.1 to fix astropy deprecation warnings.
- Fixed some bugs related to handling of the
UVCal.time_range
parameter inUVCal.__add__
and other methods. Ensured that it is a list whenUVCal.future_array_shapes
is False. - Fixed a bug in
UVCal.__add__
when data are sorted differently along any axis or interleaved along an axis. - Fixed a bug in reading MS files with non-UTC time scales.
- Fixed some bugs in UVData, UVFlag and UVCal
__add__
andselect
methods for objects with only one spectral window andflex_spw_id_array
defined. Also fixed some handling in those methods on UVFlag for objects with multiple spectral windows. - Fix a bug when reading FHD files into UVData objects when the layout file was not passed but the antenna_positions are in the known telescopes.
- Fixed a bug where the beamfits partial read parameters were not available in the
UVBeam.from_file
class method and were not passed properly when reading multiple files. - Fixed a bug where objects created from scratch with the old phase attributes weren't
properly getting converted to the new
phase_center_catalog
representation. That conversion now happens in the check so will happen before anywrite
. There could still be errors if some methods are called before a check is done (or if the check is turned off). - Fix a bug in how
frame_pa
was calculated inphase
and_set_app_coords_helper
for multi_phase_center objects (it was using the oldphase_center_frame
attribute). - Fix a bug where trying to select lsts or lst_ranges on read didn't work for some file types.
- Severe performance hit when calling
polnum2str
and its variants for many baselines.
- Reading files into objects without setting
use_future_array_shapes
now results in deprecation warnings. - The utility functions
phase_uvw
andunphase_uvw
associated with the deprecated old phasing method. - The
flex_spw_id_array
will be required on all UVData and UVFlag and all non-wide-band UVCal objects in version 3.0. - Deprecated the
interpolation_function
attribute on UVBeams. - Deprecated the older phase attributes (
phase_type
,phase_center_ra
,phase_center_dec
,phase_center_frame
,phase_center_epoch
,object_name
) in favor of thephase_center_catalog
. The older phase attributes will be removed in version 3.0. - The
lst_from_uv
function in the uvflag module.
- Removed deprecated handling for UVCal objects without
telescope_location
,antenna_positions
andlst_array
parameters defined.
- Updated how wheels for PyPI are built.
- A frame attribute to the telescope_location parameters on UVData, UVCal and UVFlag objects to support observatories on the moon (or anywhere not on earth).
- A bug in
UVData.read_uvh5
wheremulti_phase_center
was being used to imply the existence of thephase_center_catalog
header item, rather than checking for the presence of that item. This conflicted with the uvh5 memo.
- A new subclass of UVParameter for SkyCoord objects, needed in pyradiosky.
- Added
UVData.convert_to_flex_pol
method to enable writing uvh5 files with the polarization axis changing slowest as desired by HERA. Also updatedUVData.remove_flex_pol
to properly undo theUVData.convert_to_flex_pol
operation. - Added an option to
UVData.read
andUVData.from_file
methods when reading uvh5 type files to callremove_flex_pol
as part of the read. This is defaulted toTrue
so that when HERA data is written with flex_pol the change will not disrupt users, but other users of uvh5 may want to set it toFalse
. - Added support to make bitshuffle compression easier when writing UVData objects to uvh5 files
- Support for partial reads on beamfits files. This can be done on frequency, azimuth or zenith angle (if beam is in az/za coordinates).
- Logging statements (INFO level) in
UVData.check
.
- Significantly increased uvh5 reading and writing speed.
- Added
use_future_array_shapes
method to UVBeam to allow users to convert to using the future array shapes now, with support throughout UVBeam methods and related utility functions. Also addeduse_current_array_shapes
to revert to the standard shapes. - Major performance improvement in the
UVBeam.to_healpix
method. - Performance improvement when doing a
UVData.select
using thebls
parameter (~70% improvement in the_select_preprocess
function)
- A bug in
UVData.check
whencheck_autos
is True and the object is a flex pol object.
- Added
use_future_array_shapes
method to UVFlag to allow users to convert to using the future array shapes now, with support throughout UVFlag methods and related utility functions. Also added ause_future_array_shapes
parameter to UVFlag's__init__
,read
,from_uvdata
andfrom_uvcal
methods to allow the conversion to be done at the time the object is made. Addeduse_current_array_shapes
to revert to the standard shapes. - Baseline number calculations for antenna numbers > 2047.
- New uvh5 version 1.1 definition in memo and code to use the future phasing info (i.e. the phase_center_catalog) rather than the older parameters.
- Added the
normalize_by_autos
method toUVData
, which allows one to normalize cross-correlations (i.e., covert values from arbitrary scale to correlation coefficients) if auto-correlations are present in the data. - Added the
write
method toMirParser
, which allows forMirParser
to write out Mir-formatted data to disk. - Added the
select
method toMirParser
, which provides a much simpler interface for selecting subsets of data to load. - Added the
rechunk
method toMirParser
, which allows data to be spectrally averaged (either after already loading or to be averaged "on-the-fly" when data are read from disk). - Added the
redoppler_data
method toMirParser
, which enables frequency shifting of the data. - Added new
__add__
and__iadd__
methods toMirParser
for combining data sets. - Added new functionality to
UVBeam.check
to verify that power beams for the auto polarizations (and pstokes) are real-only, along with an option to force them to be real-only if non-zero imaginary components are detected. - Coarse band correction for mwax to
mwa_corr_fits
- Reordering methods for UVCal:
reorder_antennas
,reorder_freqs
,reorder_times
andreorder_jones
.
- Always write out uvh5 version 1.0 or greater files (which use the future array shapes) with
UVData.write_uvh5
. - Use "unprojected" rather than "unphased" for the cat_type in the phase_center_catalog to describe zenith driftscan data that have no w-projections applied.
- The
MirParser
class has been significantly overhauled, which significantly reduces memory usage and improves processing speed. - Updated minimum dependency versions: numpy>=1.19, scipy>=1.3, optional dependencies: python-casacore>=3.3, pyyaml>=5.1, astropy-healpix>=0.6
UVBase
object now require that individual attribute names match that given inUVParameter._name
.UVData.fix_phase
now raises a warning when called.- Changed
UVData.write_uvfits
to allow for one to write out datasets in UVFITS format without "spoofing" (via settingspoof_nonessential=True
) UVFITS-specific values. - Methods that ensure that attributes are set up properly on
UVBeam
are now called in theUVBeam.check
method, making it easier to fill in UVBeam objects from scratch. - Changed uvfits reading and writing to no longer subtract/add 1 to the antenna numbers. The subtraction/addition of 1 was initially done because FITS uses one-based indexing but antenna numbers are physical numbers not indices, so the numbers should not be adjusted.
- Changed the way baseline numbers are calculated from antenna numbers to conform to the standard in uvfits and miriad. The inconsistency with those files was noticed when we corrected the handling of antenna numbers in uvfits.
- The
spw_order
parameter for theUVData.reorder_freqs
method now accepts an index array rather than an array of spw numbers, making it match the other reorder methods. - Updated the astropy requirement to >= 5.0.4
- Dropped support for python 3.7
- Store the phase_center_catalog as a set of nested datasets in uvh5 files rather than as a json blob.
- A bug in
UVData.__add__
where flexible spectral window datasets were not combined correctly if they contained overlapping frequencies in different spectral windows. - A bug in
UVData.print_phase_center_info
that occasionally resulted in incorrect values being reported for RA/Az/Longitudinal coordinates. - A bug in
UVData.select
that could causeUVData.check
to fail ifUVData._scan_number_array
was set. - A bug in
UVBeam.select
where after selecting down to only auto polarization power beams theUVBeam.data_array
remained complex instead of real. - A bug in
UVBeam.__add__
where adding an object with cross pol power beams to an object with only auto pol power beams could result in loss of the imaginary part of the cross pol power beams. - Testing of piping keywords through
read
bytest_mwa_corr_fits
. - Incorrect piping of
flag_dc_offset
keyword. mwa_corr_fits
handing of theBSCALE
keyword in gpubox files.
- The "unphased" cat_type in phase_center_catalog.
- Parameter
check_azza_domain
inUVBeam.interp()
, which can turn off the (relatively expensive) check on input az/za values.
- The
UVCal.initialize_from_uvdata
method to initialize a UVCal object using metadata from a UVData object. - An option to
UVData.reorder_blts
to sort the autos before the crosses. - Added new functionality to
UVData.check
to verify that auto-correlations are real-only, along with an option to force them to be real-only if non-zero imaginary components are detected.
UVFlag.to_baseline()
uses internal time tolerances (both rtol and atol) stored in the UVParameter_time_array.tols
to check whether times in the UVFlag object and the input UVFlag/UVData object match.
- The
with_conjugates
keyword in favor of theinclude_conjugates
keyword in theutils.get_baseline_redundancies
function.
- A new attribute to
UVData
calledflex_spw_polarization_array
, of type=int and shape=(Nspws
,), which allows for individual spectral windows to carry different polarization data. - Added the
_make_flex_pol
andremove_flex_pol
methods toUVData
, which allows for one to convert a standardUVData
object to one with "flexible-polarization". - Added a new method to
MirParser
called_apply_tsys
, which will convert MIR visibility data from correlation coefficients to pseudo-Jy. - Added a method to
Mir
called_init_from_mir_parser
, which allows for one to pass aMirParser
object to be converted into a UVData object (rather than reading from a file on disk). - Added the
flex_spw
attribute to theUVCal
class, which can be set to True by using the new_set_flex_spw
method. - Added the optional
flex_spw_id_array
attribute to UVCal class, of type=int and shape=(Nfreqs
,), which indexes individual channels along the frequency axis tospw_array
. - Added
use_future_array_shapes
method to UVCal to allow users to convert to using the future array shapes now, with support throughout UVCal methods and related utility functions. Also addeduse_current_array_shapes
to revert to the standard shapes. - Added support for wide-band gain calibrations via the
wide_band
attribute on theUVCal
class, which can be set using the new_set_wide_band
method. - Added
time_range
,lsts
, andlst_range
kwargs from UVH5.write_uvh5_part() to UVData.write_uvh5_part(). - Added
time_range
,lsts
, andlst_range
kwargs from UVH5.write_uvh5_part() to UVData.write_uvh5_part().
- General performance improvements in the
read_mir
method. - Enabled
read_mir
to read in dual- and full-polarization data.
- Fixed a bug where
UVData.compress_by_redundancy
sometimes produced incorrectly conjugated visibilities. - Fixed a bug where
lsts
andlst_range
got ignored when doing partial i/o with multiple files.
- Support for reading MWAX correlator fits files.
- Added a warning when using
UVData.write_uvfits
when thevis_units
attribute is anything other thanJy
. - Added a warning when using
UVData.write_uvfits
if a UVData object has > 256 antennas, recommending use ofUVData.write_ms
if intending to import the data into CASA. - Reading writing of scan numbers for MS files as
UVData.scan_number_array
. - Grouping of contiguous integrations for a phase center into "scan numbers" in
UVData._set_scan_numbers
. This grouping definesUVData.scan_number_array
when not originally present in the data (e.g. reading in a non-MS file) and is used when writing to an MS file. - Added a
filename
attribute to UVCal, UVBeam and UVFlag objects. - Support for reading MWAX/birli uvfits files.
- Flexible spectral windows to
read_mwa_corr_fits
.
- Changed the
UVData.write_ms
to reduce processing time and memory footprint. - Changed the order in which data are written in Measurement Set format via the method
UVData.write_ms
, where data are now first grouped together byscan_number_array
. - Changed the defaults by which the
UVData.antenna_names
attribute is set when reading measurement sets (used to default to station names, now uses antenna names if available and the MS file was not written using the CASAimportuvfits
routine). - Fixed a bug where
UVData.x_orientation
andUVData.vis_units
were not being written to / read from measurement sets. - Updated the astroquery requirement to >= 0.4.4, due to changes in the API handling for calls to JPL Horizons.
- Assumes uvfits files are in ITRF frame unless explicitly stated otherwise. Consistent with AIPS 117.
- Improved readability, functionality, and memory usage in
read_mwa_corr_fits
.
- A bug in writing calfits files when the optional
time_range
parameter is not set. - A bug where the
TIME_CENTROID
field was not being filled in data sets written byUVData.write_ms
, which caused odd behavior in some CASA routines (e.g.,gaincal
). - A bug in reading in uvfits files with baseline coordinates that have suffixes of '---SIN' or '---NCP' which are allowed in uvfits files.
- A bug that could cause some routines in CASA to fail when using data sets written by
UVData.write_ms
. - A bug that could have resulted in
UVData.__add__
combining objects together incorrectly when containing overlapping time-baselines/polarization/frequency channels together that were ordered differently for the twoUVData
objects. - A bug that prevented
extra_keywords
keys with a value ofNone
from being saved to UVH5 files. - A bug that resulted in the wrong expected shapes for data-like arrays when metadata only UVData objects were set to use future array shapes.
- A bug that caused an error when writing non-double antenna diameters to measurement set files
- Added a warning in
utils.uvcalibrate
when uvdata x_orientation is not set. - Fixed a bug in
UVBeam.efield_to_power
when there is only one feed.
- Fixed a bug in converting non-crosspol power beams to real in
UVBeam.efield_to_power
.
- A bug that resulted in coordinate queries via for solar-system objects
utils.lookup_jplhorizons
to throw an error. The workaround temporarily requires the precision for ephemerides to be limited to 0.05 arcseconds. - A bug that resulted in a potential error when attempting to write out multi-phase-center data sets into the UVH5 file format.
- Added the
UVData.write_ms
method, a UVData measurement set writer. - Added an option to the
UVData.check
method to flip the uvws and conjugate the data_array to align with theUVData
convention on baseline orientation. - Added the
flip_gain_conj
option toutils.uvcalibrate
to use the opposite gain conjugation convention in calibration. - Support for selecting on polarization strings on
UVData
,UVCal
,UVBeam
andUVFlag
. from_file
class function on UVData and UVBeam. Allows users to instantiate an object from a file directly from each Class.- Generic 'read' function for UVBeam objects.
- UVData methods
set_data
,set_flags
, andset_nsamples
which allow the user to update values in the corresponding arrays. - pyERFA was added as an explicit dependency (previously it was implicit because it is an astropy dependency).
- A bug in the way times were written to uvfits files that could cause a loss of precision.
- A bug that allowed the check to pass on
UVData
,UVCal
andUVFlag
objects if theflag_array
contained integers instead of booleans. - A bug in equality testing for
UVParameter
objects with strict types when the parameters contain arrays. - Provide a more useful error message if
UVCal.read_fhd_cal
is called withread_data=False
and a settings file is not provided. - Fixed a bug in
UVCal.read_fhd_cal
where the reader crashed on a metadata only read if the settings file had an empty field for the diffuse model. - Fixed a bug in
UVBeam.efield_to_power
where a thedata_array
remained complex rather than real because the tolerance was too low in the numpyreal_if_close
call.
- The
__eq__
method on UVBase objects (and subclasses) now supports anallowed_failures
keyword for parameters which are allowed to be unequal without failing the check. This includes thefilename
parameter ofUVData
objects by default.
- Added
filename
attribute for UVData objects. - Added support for multiple sources/phase centers in a single UVData object, which can
be enabled using the
_set_multi_phase_center
method. Using this method will set the new attributemulti_phase_center
toTrue
(otherwise set toFalse
.). - Added the optional attribute
Nphase
, of typeint
and required ifmulti_phase_center=True
, which records the number of different phase centers recorded in the UVData object. - Added the optional attribute
phase_center_catalog
, a dictionary of dictionaries that is required ifmulti_phase_center=True
, which records various meta-information about the different phase centers used within the UVData object. The keys of individual entries of phase_center_catalog are strings representing the names of the individual phase centers. These keys are matched to individual dictionaries which store various information about the phase centers, including coordinate information (e.g.,cat_lat
,cat_lon
,cat_frame
,cat_epoch
) and a unique identification number (cat_id
, typeint
). Catalog information can be printed to the terminal using the methodprint_phase_center_info
. - Added the optional attribute
phase_center_id_array
, of typeint
and shape(Nblts,)
required ifmulti_phase_center=True
, which records which phase center the individual baseline-time record is phased to. The individual values of are matched tocat_id
from the individual dictionary entries ofphase_center_catalog
. - Added the attributes
phase_center_app_ra
andphase_center_app_dec
, which records he topocentric apparent positions of the phase center as seen from the positiontelescope_location
. Both are of typefloat
, shape(Nblts,)
. - Added the attribute
phase_center_frame_pa
, of typefloat
, shape(Nblts,)
, which encodes the position angle between the topocentric apparent frame and the celestial frame of phase center coordinates, as recorded inphase_center_frame
(e.g., FK5, ICRS). UV-coordinates are rotated byphase_center_frame_pa
, such that - Added a switch called
ignore_name
to__add__
which allows one to ignore whether theobject_name
attribute matches when attempting to combine objects (default isignore_name=False
, which will cause an error to be thrown ifobject_name
does not match). This parameter is also available with theread
method, when reading in multiple files. - Added a switch called
make_multi_phase
to the__add__
method, which will convert the data to a multi-phase-center-enabled dataset, allowing for UVData objects with different phase centers to be combined. This parameter is also available with theread
method, when reading in multiple files. - Added the
fix_phase
method, which converts UVData objects from the "old" phasing framework to the "new' phasing framework. - Added support for looking up ephemeris information of solar system objects from
JPL-Horizons via the optionally required
astroquery
packaged. - Added option to select based on LST or an LST range in UVData objects.
- Added
get_lsts
method on UVData objects for retrieving LST corresponding to data.
utils.uvcalibrate
will error rather than warn if some of the newly added checks do not pass, including if antenna names do not match between the objects.- Phasing methods (e.g.,
UVData.phase
,UVData.unphase_to_drift
,UVData.set_uvws_from_antenna_positions
) have undergone a significant revision, which change the way in UVW coordinates are calculated. Specifically, these methods will now use the topocentric apparent positions (stored in the attributesphase_center_app_ra
,phase_center_app_dec
). As a result, phasing of data and calculating of uvw-coordinates are significantly faster. phase
method now supports phase centers of different types, specified by thecat_type
parameter. Supported types include "sidereal" (default; fixed position in RA/Dec), "ephem" (position in RA/Dec which moves with time), and "driftscan" (fixed position in Az/El, NOT the same asphase_type
=“drift”). The "ephem" and "driftscan" types can only be used whenmulti_phase_center=True
.phase
now uses antenna positions for deriving uvw-coordinates by default.- Decreased the standard tolerances for angle-based parameters to 1 mas (was 10 mas).
- Updated the astropy requirement to >= 4.2.1
- Unknown phasing types (i.e.,
phase_type="unknown"
) are no longer supported,read
(and the associated file-reading methods) will defaultphase_type
to "drift" when unable to definitively ascertain the phasing type.
- A bug in
aipy_extracts.pipe
that caused an error when usingraw=True
. - A bug in
aipy_extracts.write
which caused an error when using certain masked arrays. - Phasing accuracy/UVW coordinate calculation no longer limited to 2 cm on a 3 km baseline (i.e.,~1 part in 1e5; new accuracy is better than 1 part in 1e8).
- A bug in reading DSB data from Mir filetypes, that resulted in visibilities being matched to the wrong baselines.
- A bug in
UVData.phase_to_time
which prevented users from rephasing phased objects. - A bug in
UVBeam.read_cst_beam
when specifying beams with a single file using a yaml.
- The UVBeam methods
set_cs_params
,set_efield
,set_power
,set_simple
, andset_phased_array
have been removed. - The UVCal methods
set_gain
,set_delay
,set_unknown_cal_type
,set_sky
, andset_redundant
have been removed. - The UVData methods
set_phased
andset_drift
have been removed. - The functions
utils._str_to_bytes
andutils._bytes_to_str
have been removed. - The
flag_missing
keyword of theutils.uvcalibrate
function has been removed.
- Added
use_future_array_shapes
method to allow users to convert to using the future array shapes now, with support throughout UVData methods and related utility functions. Also addeduse_current_array_shapes
to revert to the standard shapes. - Added versioning to UVH5 files. Files with future array shapes are version 1.0, files with the current shapes are are version 0.1.
- Added
UVData.reorder_freqs
method to reorder the data along the frequency axis. - Better re-initialization of UVParameters on UVBase objects when using pickle. Improves compatibility of UVBase objects with MPI.
- Added option to apply a Van Vleck correction to mwa_corr_fits files.
- A bug in mwa_corr_fits that resulted in an error when using
memap=true
while reading files containing the 'BSCALE' keyword. - A bug in mwa_corr_fits that didn't properly test against 2015 data, and so didn't
catch the error when using
memap=true
while reading files containing the 'BSCALE' keyword. - A bug in mwa_corr_fits that didn't round start_flag to the nearest multiple of integration time when using goodtime, resulting in an error when reading data taken at 2 seconds.
- A bug in reading in the MWA beam that caused the beam to be rotated by 90 degrees.
- Added
propagate_coarse_flags
option toread_mwa_corr_fits
, as well as new flagging defaults. - Added support for
extra_keywords
in UVFlag construction from UVData objects (and paths)
- Improved memory usage in reading MWA correlator fits files.
- Speed improvement in redundant group finder.
- Fixed a bug in
compress_by_redundancy
withmethod='average'
that resulted in more unique times on the compressed object than on the original object. - Fixed bug causing MemoryError when finding redundant baselines for over 100 antennas.
- Added support for
telescope_location
,antenna_positions
andlst_array
in UVCal objects and file types. - Added support for a
metadata_only
mode in UVCal, including options to only read the metadata when reading in calibration files. - Added a
copy
method for UVCal objects. - Added
antenna_positions
,antenna_names
andantenna_numbers
as optional metadata to our known telescopes list and added HERA antenna position information. - Added handling for
extra_keywords
in UVFlag objects, including read/write to/from files. - Added handling for
object_name
andextra_keywords
tosum_vis
anddiff_vis
methods and added theoverride_params
option to override other parameters.
- Changed to use Astropy sites for telescope locations when avaliable. This results in a small change for our known position for the MWA.
- Modified
UVData.read
to do faster concatenation of files, changed the interface toUVData.fast_concat
to allow lists ofUVData
objects to be passed in.
- Fixed a bug where telescope positions from MWA uvfits files created by Cotter were not identified as being in the ITRF frame because of a missing FITS keyword.
- Fixed a bug where
antenna_positions
from FHD files were interpreted as being in the relative ECEF frame rather than the rotated ECEF frame - Fixed a bug where the
lst_array
was not updated incompress_by_redundancy
when usingmethod='average'
. - Fixed an undefined parameter bug when interpolating beams that do not cover the sky.
- Fixed an indexing error bug in
compress_by_redundancy
when usingmethod='average'
.
- UVCal objects now require
telescope_location
,antenna_positions
andlst_array
parameters. Not setting them will cause an error starting in pyuvdata 2.3.
- Added support for multiple spectral windows for UVH5, UVFITS, MIR, and MIRIAD files.
- Added the
flex_spw
attribute to theUVData
class, which can be set to True by using the new_set_flex_spw
method. - Added the optional
flex_spw_id_array
attribute, of type=int and shape=(Nfreqs
,), which indexes individual channels along the frequency axis tospw_array
. - Adjustment to digital gain removal from mwa_corr_fits files to account for a division by 64 due to a bit selection shift.
- Options to remove coarse band shape and digital gains from mwa_corr_fits files.
- Support for cotter flags in mwa_corr_fits files.
- When
flex_spw=True
, thechannel_width
attribute of UVData is expected to be an array of type=float and shape=(Nfreqs
,). Individual channels are allowed to have different channel widths. - Changed
freq_array
to be of shape (1,Nfreqs
) (formerley (Nspws
,Nfreqs
)). - Changed
data_array
,flag_array
,nsample_array
to be of shape (Nblts
, 1,Nfreqs
,Npols
) (formerly (Nblts
,Nspws
,Nfreqs
,Npols
)). - Changed UVH5 metadata byte conversion method from tobytes() to bytes()
- Added read-only support for MIR files, adding the
Mir
class toUVData
. - Added the
MirParser
class, which allows for python access to MIR files. - A new
check_warnings
method in our test module that behaves likepytest.warns
but adds the ability to check for multiple warnings. - The 'ant_str' option when making selects on UVFlag objects
- A check that the uvws match the antenna positions, as part of the acceptability checking.
- Updated the numpy requirement to >= 1.18
- Moved
parse_ants
toutils.py
to allow any UVBased object the potential to use it.
- The
checkWarnings
function in our test module, use the newcheck_warnings
function instead.
UVFlag
can now takePosixPath
objects as input.- An option to average rather than select a single baseline in
UVData.compress_by_redundancy
, controlled by themethod
keyword. nsample_array_dtype
keyword forUVData.read
, which sets the datatype for the nsample_array. Currently only used for mwa_corr_fits files.utils
cython module to speed up some common utility functions.- Added support for installing the package on Windows.
- 'background_lsts' keyword in
uvdata.read
to calculate lsts in the background while reading. Defaults to True. background
keyword inuvdata.set_lsts_from_time_array
to allow lst calculation in a background thread.read_data
keyword to read_mwa_corr_fits, allows from metadata only reads of MWA correlator fits files.propagate_flags
keyword forUVData.frequency_average
which flags averaged samples if any contributing samples were flagged
- Nants data calculation changed to use numpy functions for small speed up.
- Changed
input
variable toindata
in UVFlag to avoid shadowing python builtin. select
now also accepts a list of baseline indices for thebls
parameter.data_array_dtype
keyword forUVData.read
is now also respected by mwa_corr_fits files. Previously it was only used by uvh5 files.- The
time_range
parameter on UVCal is no longer required and it is suggested to only set it if the Ntimes axis is length one. - FHD now supports metadata only reads.
- Updated formatting and added more explicit typing in corr_fits.pyx
- Updated Nants calculations for speed up.
- updated HERA telescope location to match the HERA defined center of array.
utils.uvcalibrate
now incorporates many more consistency checks between the uvcal and uvdata object. The new keywordstime_check
andant_check
were added to control some of these checks.
- The UVBeam methods
set_cs_params
,set_efield
,set_power
,set_simple
, andset_phased_array
have been made private. The public methods will be removed in version 2.2. - The UVCal methods
set_gain
,set_delay
,set_unknown_cal_type
,set_sky
, andset_redundant
have been made private. The public methods will be removed in version 2.2. utils.uvcalibrate
will error rather than warn if some of the newly added checks do not pass, including if antenna names do not match between the objects, starting in version 2.2. In addition, theflag_missing
keyword is deprecated and will be removed in version 2.2.
- UVFlag.add now properly concatenates all existing data-like parameters of the object, including optional ones.
- A bug in
UVData.downsample_in_time
where the data were not being properly weighted by the nsample array and the nsample_array was not being properly weighted by the integration times. - A bug in
UVData.downsample_in_time
that lead to duplicated data on the final object if a baseline had varying integration times and some integration times were greater than or equal to the requested minimum integration time.
- New
_corr_fits
C extension for performing the index and conjugation mapping calculation for read_mwa_corr_fits copy
method to UVCal and UVBeam objectschunks
toUVH5.write_uvh5
andUVData.write_uvh5
for HDF5 dataset chunking.multidim_index
toUVH5.read_uvh5
andUVData.read
for multidimensional slicing into HDF5 datasets
- Various serialized calculations in uvdata.py, uvbeam.py, and utils.py to be vectorized with numpy
- Miriad interface was re-written from a hand written CPython interface to a Cython interface which dynamically creates the CPython during package setup/build.
- uvfits, calfits and beamfits files should now be read faster due to a simplified handling of fits header objects, especially when the history is very long.
- The UVData methods
set_drift
,set_phased
, andset_unknown_phase_type
have been made private. The public methods will be removed in version 2.2.
- UVData allows conjugation of metadata only objects.
- Handling of strings in UVFlag files has been made more widely compatible.
- Option to provide parameters for RectBivariatespline through interp
weights_square_array
(optional) parameter on UVFlag - stores sum of squares of weights when converting to waterfallfrequency_average
method on UVData to average data along the frequency axis.
- A bug in
select
where baseline-polarization tuples would not be conjugated correctly inUVData
andUVFlag
, and select on read for Miriad files. metafits_ppds.fits
files can now be passed tomwa_corr_fits.read
without throwing an error.- UVParameters that are array_like and have NaNs are now properly identified as equal if the NaNs are in the same locations and all non-NaN entries are equal.
- A bug in
mwa_corr_fits.read
in filling flag and nsample arrays. - A bug in
UVData.downsample_in_time
in calculating the number of new blts.
- All references to Python 2 removed from codebase.
- Routine flagging of MWA coarse band edges and center channels, as well as beginning and end integrations. Only done during
read_mwa_corr_fits
. - Added a return_weights_square option to
utils.collapse
function, which returns the sum of the squares of the weights.
- Previously deprecated code marked for removal in version > 1.5:
- reading multiple files with file format specific read functions (e.g. read_mirad). Multi-file reads can only be performed with
uvdata.read
read_metadata
keyword in various specific read functions (e.g.read_miriad
)metadata_only
keyword inselect
andget_redundancies
uvdata.miriad.read_miriad_metadata
phase_center
keyword inuvdata.read_mwa_corr_fits
phase_data
keyword inuvdata.read_mwa_corr_fits
uvdata.get_antenna_redundancies
uvdata.get_baseline_redundancies
uvdata.uvfits.read_uvfits_metadata
uvdata.uvfits.read_uvfits_data
- Support for reading the MWA full embedded element beam into a UVBeam object.
- New
time_range
keyword toselect
so exact times don't need to be specified (also added toread
methods for select on read). - Support for rephasing phased data including on
read
,__add__
andfast_concat
so that files with different phasing can be read in together. sum_vis
anddiff_vis
for summing or differencing visibilities in the data_array.read_mwa_corr_fits
for reading in MWA correlator gpubox files and applying cable corrections.eq_coeffs
for storing equalization coefficients andremove_eq_coeffs
for removing them.utils.apply_uvflag
for applying UVFlag objects to UVData objects
- Fixed
utils.uvcalibrate
to handlex_orientation
attribute - Arrays of file names can now be passed to
read
(not just lists). - run_check is no longer turned off in
read_uvfits
andread_uvh5
whenread_data
is False. - Redundancy finder will now error if any baselines appear in multiple groups.
- A bug in
UVCal
objects that prevented them from properly getting data withee
/nn
-style polarizations. - a bug in
UVFlag
wherex_orientation
was not set during initialization. - A bug in
UVCal.read_fhd_cal
that caused calibration solutions to be approximately doubled. - A bug in UVFlag where polarization array states were not updated when using
force_pol
keyword into_antenna
andto_baseline
- A bug in UVFlag.to_baseline() where force_pol kwarg did not work for UVData Npols > 1
UVData.read_uvfits
no longer breaks if there are non-ascii bytes in antenna names (which CASA sometimes writes).
- Reading in multiple files (or file sets) using file-type specific read methods (e.g.
read_uvfits
) in favor of the genericread
method. - The
phase_center
andphase_data
keywords toread_mwa_corr_fits
in favor ofphase_to_pointing_center
and thephase_center_radec
keyword in the genericread
method. - Support for reading only the header (not all the metadata) of uvfits files.
- The
read_uvfits_metadata
andread_uvfits_data
methods on the UVFITS object. - The
read_miriad_metadata
method on the Mirad object.
copy
method forUVData
, which can optionally make a copy of just metadataupsample_in_time
,downsample_in_time
, andresample_in_time
methods onUVData
objectsutils.uvcalibrate(.., undo=True)
kwarg for undo-ing a calibration.utils.uvcalibrate
updatesUVData.vis_units
ifUVCal.gain_scale
is set.UVCal.gain_scale
non-required attribute- UVData.get_redundancies method to replace old get_baseline_redundancies and get_antenna_redundancies.
- option for
UVBeam.interp
to return a new beam object. UVFlag
information on Read The Docs
UVData.phase_to_time
now accepts a float as an input. Assumes float represents a JD.- Added optional acceptability check for
utils.LatLonAlt_from_XYZ
- Use
astropy_healpix
rather thanhealpy
for HEALPix functionality becausehealpy
has a GPL license which is incompatible with ours. h5py
is now a required package instead of an optional one.- Phasing now supports metadata only
UVData
objects - utils.get_baseline_redundancies uses scipy pdist functions instead of for loops (faster)
- UVData.get_antenna_redundancies will no longer automatically conjugate baselines.
- UVData.get_baseline_redundancies and UVData.get_antenna_redundancies have been combined.
UVFlag
inherits fromUVBase
object.UVFlag
objects can now convert from antenna type to baseline typeUVFlag
objects can now be initialized without inputs
- A bug in UVBeam._interp_freq where kind parameter was not passed for real-only beams
- A bug in get_antenna_redundancies for nearly N-S baselines.
- A bug where
conj_pol
could not handle cardinal direction polarizations. - A bug that gave the wrong error message when calling
UVData.phase_to_time
without an Astropy Time object.
- UVData.get_baseline_redundancies and UVData.get_antenna_redundancies will be deprecated in version 1.6.1
metadata_only
property onUVData
to automatically detect if data-like arrays are present- support for combining metadata only objects and reading in multiple files as metadata only
utils.uvcalibrate
flag propagation bug fixUVCal.ant2ind
indexing bug fixUVCal.get_*
methods for accessing data arrays with antenna-polarization keysutils.uvcalibrate
for automated calibration ofUVData
byUVCal
- Fixed a bug in select that caused bls and antenna_names/numbers to be or'ed rather than and'ed together.
- Fixed a bug where
baseline_to_antnums
could accept a numpy array as input but not other array_like objects.
- removed
new_object
keyword fromUVBeam._interp_freq
in favor of new functionality inUVBeam.interp
- Option in UVBase.check() to ignore whether required parameters are set.
- made an option to not save the
lst_array
to uvfits files. conjugate_bls
option toUVData.get_antenna_redundancies
UVData.conjugate_bls
method to conjugate baselines to get the desired baseline directions.UVData.reorder_blts
method to reorder the data along the blt axis (and optionally also conjugate baselines), and a newblt_order
optional parameter on UVData objects to track the ordering (including through read/writes).lst_array
is now saved to UVFITS files (even though it's not a standard parameter) so that it doesn't have to be recalculated
- Fixed init logic in UVFlag.
- Fixed a bug in how FHD uvw vectors were oriented (and visibilities were conjugated)
- Fixed a bug in
UVData.inflate_by_redundancy
when Nblts is not equal to Nbls * Ntimes. - Fixed a bug in UVData when reading in an FHD file with a single time integration.
- Fixed a bug in how the longitudinal branch cut was handled in beam interpolation
- Changed the way interpolation splines are saved in UVBeam to fix errors related to polarization selections.
- Python 3:
np.string_
call now useskeepdims=True
to guard againstantenna_names
being cast as an array.
- Testing framework changed from
nose
topytest
. uvdata.set_lsts_from_time_array
only calculates lsts for unique elements intime_array
.
- Optional
x_orientation
parameter to utils functions polstr2num, polnum2str, jstr2num and jnum2str to allow for E/N based polarization strings (rather than just x/y based ones) - New optional
x_orientation
parameter on UVBeam (paralleling UVData and UVCal), with read/write support in beamfits - Added
x_orientation
as an optional parameter in read_cst_beam and in cst settings yaml files. - All str2num or num2str calls on UVData and UVBeam pass the object's x_orientation
- New
UVData.fast_concat
method to allow fast concatenation of UVData objects (or files) along a particular axis. - Added preliminary
UVFlag
module from hera_qm to pyuvdata. Will eventually promote toUVBase
object, but for now this is undocumented functionality.
- Defined 'east' and 'north' as the allowed 'x_orientation' values in UVData and UVCal, Backwards compatiblity support exists for 'E' and 'N' values
UVData.order_pols
method in favor ofUVData.reorder_pols
.
- Building pyuvdata on macOS now targets minimum macOS 10.9 if run on macOS 10.9 or above
- Possible bug where
check_variables
dictionary can change size duringread_miriad
call
- Added
add_to_history
kwarg to UVH5.write_uvh5_part _healpix_interp_bilinear
as a new interpolation method inUVBeam
freq_interpolation_kind
added as an attribute toUVBeam
tol
added as keyword argument toUVBeam._interp_freq
which allows for a fast return ofdata_array
slice if nearest-neighbor frequencies are all within the distance tolerance.polarizations
added as keyword argument toUVBeam
interpolation methods.- Support for a yaml settings file to collect and propagate metadata for CST beam files.
UVBeam._interp_freq
returns bothinterp_data
andinterp_bandpass
, instead of just the former.
- Combining overlapping data along multiple axes (most common when reading in multiple files) no longer errors.
keep_all_metadata
keyword for optionally discarding unused metadata when performing a select operation.
- Extends
run_acceptability_check
for UVH5 metadata incheck_header
function.
- Antenna numbering bug in redundancy methods. It wasn't using the correct antenna numbers to make baseline indices.
- Redundancy code returns one group if all baselines are redundant. Previously returned each baseline as a separate group
- Redundancy code finds unique baselines along baseline_array without assuming Nblts = Nbls * Ntimes. Previously assumed Nblts = Nbls * Ntimes and attempted to slice array.
- "inflate_by_redundancy" method errored when phase_type == phased, due to _set_u_positive using phased uvw coordinates. It now uses ENU frame uvw coordinates.
- Methods on UVData objects to compress/inflate data by redundant baselines.
- Convenience functions on UVData for finding redundant baselines (calling the corresponding utils functions)
- memo describing the UVH5 format
- read/write support for uvh5 files with integer datatypes for visibilities
- Option to only do the select on the metadata. This is useful for partially defined objects as in pyuvsim setup or after reading only the metadata from a file.
- support for python3.7
- UVdata.get_ENU_antpos() now defaults to using the telescope_location as the center rather than the median antenna position.
- UVBeam.efield_to_pstokes() no longer restricted to healpix coordinates
- latitude and longitude in uvh5 files are written in degrees instead of radians.
- Fixes a bug in redundancy methods for when there are no redundant baselines.
_key2inds
now properly reorders polarization axis for conjugated visibilities. This also effects theget_data
function.- long strings are saved correctly in miriad files from python3
- option to save splines for reuse in UVBeam.interp function
- improve string handling for uvh5 files
- changed top-level import structure to exclude file-specific class (e.g.
UVFITS
,CALFITS
) and base classes (UVBase
,UVParameter
) and to not import utility functions into the top-level namespace
- Support for UVData objects without antenna_positions. Antenna positions will be required in a future version.
- Utility functions to identify redundant baselines (either by baseline or antenna pair)
- Reading FHD layout files to get antenna positions
- Polarization dict constants and parsing functions in utils.py for mapping between polarization strings and numbers
- LST array is now optional in uvh5 files
- polarization string capitalization was unified throughout: use lower case except for Stokes parameters
- integration_time is now a vector of length NBlts instead of a scalar
- Support for FHD inputs without layout files (containing antenna positions).
- flags are always returned as a boolean array from
UVData.get_flags
- integration_time, uvw_array and lst_array are now always checked for consistency when adding UVData objects
- consistency checks on baseline lengths now uses the uvw_array tolerances
- compatibility with python 3 (retaining python 2 compatibility)
- generic
UVData.read
method for all input file types, including select on read options - partial write support for uvh5 file format
- partial read support for uvfits, miriad and uvh5 file formats (including only reading metadata and select on read)
- new uvh5 file format: an HDF5 file standard matched to UVData objects
- new method to calculate uvws from antenna positions
UVBeam.get_beam_area
andUVBeam.get_beam_sq_area
functions to calculate beam integrals, including for pseudo-Stokes beams- beam interpolation methods, to any set of points and to healpix pixel centers
- a script to renumber antennas for CASA compatiblity if there are fewer than 256 antennas but numbers higher than that
- memo describing the beam fits file format
- method to peak normalize UVBeam objects
- support for reading FHD calibrations into UVCal objects
- support for sky-based calibration metadata in UVCal and the calfits file format
- method to convert E-field beams to power beams
UVData.get_ENU_antpos
method to get ENU coordinates from antenna positions- support for
extra_keywords
on UVCal objects
- Major overhaul of phasing code, see the phasing memo in docs/references for more details
- Make all earth location coordinate conversions use same axes order (n_points, 3)
- replaced
ant_pair_nums
keyword inUVData.select
and partial read methods withbls
which supports lists like [(0,1,'xx'), (2,3,'yy')] - extracted miriad wrappers from aipy, removing aipy as a dependency
- Fixed error when combining auto-correlation only and cross-correlation only UVData objects
- Fixed error with the
UVData.select
function using a single antenna name - Fixed bugs with
UVData.get_data
,UVData.get_flags
, andUVData.get_nsamples
for conjugated baseline polarizations - Fixed a bug that caused a memory error in
UVData.write_uvfits
- Fixed bugs in interpreting the uvw direction convention for uvfits
- Fixed a bug in reading azimuth locations from CST beam files
- Fixed a bug reading in single frequency uvfits files
- Fixed a bug in reading MWA Cotter measurement sets
- Fixed units errors in cal fits files
- Fixed a serious bug where data was overwritten in the add functions if the axes were out of canonical order
- Fixed scrambled data ordering in add function
- Fixed a bug in parsing frequencies from CST file names
- support doctest in the tutorial
- utility functions for converting between Jones numbers and polarization strings
- support for
antdiam
variable in miriad files - module for reading a CST file into a UVBeam object
- utility functions for converting to rotated ECEF coordinates from ECEF coordinates
- support reading in a list of calfits files
- include
extra_keywords
to miriad files __add__
method for UVCal objectsantenna_diameters
value to UVData objectsparse_ants
method on UVData objectsx_orientation
value to UVData objects- "smart slicing" functionality to UVData objects
- convenience methods on UVData objects for easily getting data and metadata
- UVBeam object
- in-place selection for UVData objects
total_quality_array
value on UVCal objects__add__
method for UVData objects- utility functions for converting to local ENU coordinates from ECEF coordinates
convert_to_gain
method on UVCal delay-type objects- read-only support for CASA measurement sets into UVData objects
select
method on UVCal objects
- antenna names in miriad saved as strings instead of arrays of ASCII hex values
- baseline-time axis mis-ordering in add function
- handling of antenna positions in miriad and uvfits files
- selecting autocorrelation data from UVData objects
- indexing of spectral windows in calfits files
- handling of
total_quality_array
in UVCal objects when selecting a subset of data
(historical information needs to be filled in)