CHANGES.rst

0.6.6.dev (unreleased)

0.6.5 (2023-12-05)

• Fixed issue with fix from #893 not getting included in the 0.6.4 tag

0.6.4 (2023-12-05)

• Fix astropy.coordinates import #893

0.6.3 (2023-10-17)

• MNT: Clean up _astropy_init.py #892
• Find beam table from EXT name and CI fix #879

0.6.2 (2023-06-23)

• bugfix in #885,#886 correcting an error introduced in #883
• CI fixes

0.6.1 (2023-06-22)

• Fix memory issue when calling statistics() on FITS cubes. #752
• Cube mosaicing implemented #829, #844
• Varied bugfixes for numpy compatibility #858, #865, #883
• Many other small fixes & improvements that didn't get changelog entries

0.6 (2021-09-30)

• Fix convolve_to when units are in Jy/beam. Add error/warnings for operations for all operations that change the spatial resolution for Jy/beam cubes.
• Add argmax_world and argmin_world to return the argmin/max position in WCS coordinates. This is ONLY defined for independent WCS axes (e.g., spectral) #680
• Bugfix: subcube producing spatial offsets for large images #666
• Switch to using standalone casa-formats-io package for reading in CASA .image files (this was split out from spectral-cube). #684
• Make it possible to customize target_chunksize in the CASA reader. #705
• Fix support for dask.distributed. #712
• Bugfix: PhysicalType are now strings. #709

0.5 (2020-09-17)

• Bugfix: subcubes from compound regions previously did not work. #601
• Bugfix: VaryingResolutionSpectralCube.mask_channels now preserves previous mask. #620
• Refactor tests to use fixtures for accessing data instead of needing to run a script to generate test files. #598
• Refactor package infrastructure to no longer use astropy-helpers. #599
• Switch to using unified I/O infrastructure from Astropy. #600
• Bugfix: fix slicing of cubes with mask set to None. #621
• Refactor CASA I/O to use dask to access the array/mask data directly and to use only Python and Numpy to access image metadata. CASA images can now be read without CASA installed. #607, #609, #613
• Add new dask-friendly classes DaskSpectralCube and DaskVaryingResolutionSpectralCube which use dask to efficiently carry out calculations. #618
• Add new statistics method on DaskSpectralCube which allows several global statistics to be computed by loading each cube chunk only once. #663

0.4.5 (2019-11-30)

• Added support for casatools-based io in #541 and beam reading from CASA images in #543
• Add support for update_function in the joblib-based job distributor in #534
• Add tests for WCS equivalence in reprojected images in #589
• Improve error messages when CASA files are read incorrectly in #584
• fix a small bug in matplotlib figure saving in #583
• Allow for reading of beamless cubes in #582
• Add support for 2d world functions in #575 and extrema in #552
• Handle kernels defined as quantities in smoothing in #578
• Fix bug with NPOL header keyword in #576
• Convolution will be skipped if beans are equal-sized in #573
• Fix one-D sliceing with no beam in #568
• Paralellization documentation improvement in #557
• Astropy-helpers updated to 2.0.10 in #553
• Fixed some future warnings in #565
• Added a new documentation example in #548
• Added channel map making capability in #551
• Fix warnings when beam is not defined in #561
• Improvment to joblib parallelization in #564
• Add ppbeam attribute to lower-dimensional objects #549
• Handle CASA file beams in #543 and #545
• Add support for CASA reading using casatools (casa6) in #541
• Bugfix for slicing of different shapes in #532
• Fixes to yt integratino in #531
• Add unmasked_beams attribute and change many beams behaviors in #502
• Bugfix for downsampled WCS corners in #525
• Performance enhancement to world extrema in #524
• Simplify conversion from CASA coordsys to FITS-WCS #593
• Add chunked file reading for CASA .image opening #592
• Dropped python 3.5 testing in #592

0.4.4 (2019-02-20)

• Refactor all beam parameters into mix-in classes; added BaseOneDSpectrum for common functionality between OneDSpectrum and VaryingResolutionOneDSpectrum. Retain beam objects when doing arithmetic with LDOs/ (#521)
• Refactor OneDSpectrum objects to include a single beam if they were produced from a cube with a single beam to enable K<->Jy conversions (#510)
• Bugfix: fix compatibility of to_glue with latest versions of glue. (#491)
• Refactor to use regions instead of pyregion. Adds CRTF support (#488)
• Direct downsampling tools added, both in-memory and memmap (#486)

0.4.3 (2018-04-05)

• Refactor spectral smoothing tools to allow parallelized application and memory mapped output (to avoid loading cube into memory). Created apply_function_parallel_spectral to make this general. Added joblib as a dependency. (#474)
• Bugfix: Reversing a cube's spectral axis should now do something reasonable instead of unreasonable (#478)

0.4.2 (2018-02-21)

• Bugfix and enhancement: handle multiple beams using radio_beam's multiple-beams feature. This allows convolve_to to work when some beams are masked out. Also removes cube_utils.average_beams, which is now implemented directly in radio_beam (#437)
• Added a variety of stacking tools, both for stacking full velocity cubes of different lines and for stacking full spectra based on a velocity field (#446, #453, #457, #465)

0.4.1 (2017-10-17)

• Add SpectralCube.with_beam and Projection.with_beam for attaching beam objects. Raise error for position-spectral slices of VRSCs (#433)
• Raise a nicer error if no data is present in the default or selected HDU (#424)
• Check mask inputs to OneDSpectrum and add mask handling for OneDSpectrum.spectral_interpolate (#400)
• Improve exception if cube does not have two celestial and one spectral dimesnion (#425)
• Add creating a Projection from a FITS HDU (#376)
• Deprecate numpy <=1.8 because nanmedian is needed (#373)
• Add tools for masking bad beams in VaryingResolutionSpectralCubes (#373)
• Don't warn if no beam was found in a cube (#422)

0.4.0 (2016-09-06)

• Handle equal beams when convolving cubes spatially. (#356)
• Whole cube convolution & reprojection has been added, including tools to smooth spectrally and spatially to force two cubes onto an identical grid. (#313)
• Bugfix: files larger than the available memory are now readable again because spectral-cube does not encourage you to modify cubes inplace (#299)
• Cube planes with bad beams will be masked out (#298)
• Added a new cube type, VaryingResolutionSpectralCube, meant to handle CASA-produced cubes that have different beams in each channel (#292)
• Added tests for new functionality in OneDSpectrum (#277)
• Split out common functionality between SpectralCube and LowerDimensionalObject into BaseNDClass and SpectralAxisMixinClass (#274)
• added new linewidth_sigma and linewidth_fwhm methods to SpectralCube for computing linewidth maps, and make sure the documentation is clear that moment(order=2) is a variance map. (#275)
• fixed significant error when the cube WCS includes a cd matrix. This error resulted in incorrect spectral coordinate conversions (#276)

0.3.2 (2016-07-11)

• Bugfix in configuration

0.3.1 (2016-02-04)

• Preserve metadata when making projections (#250)
• bugfix: cube._data cannot be a quantity (#251)
• partial fix for ds9 import bug (#253)
• preserve WCS information in projections (#256)
• whitespace stripped from BUNIT (#257)
• bugfix: sometimes cube would be read into memory when it should not be (#259)
• more projection preservation fixes (#265)
• correct jy/beam capitalization (#267)
• convenience attribute for beam access (#268)
• fix beam reading, which would claim failure even during success (#271)

0.3.0 (2015-08-16)

• Add experimental line-finding tool using astroquery.splatalogue (#210)
• Bugfixes (211,212,217)
• Add arithmetic operations (add, subtract, divide, multiply, power) (#220). These operations will not be permitted on large cubes by default, but will require the user to specify that they are allowed using the attribute allow_huge_operations
• Implemented slicewise stddev and mean (#225)
• Bugfix: prevent a memory leak when creating a large number of Cubes (#233)
• Provide a base attribute so that tools like joblib can operate on SpectralCube s as memory maps (#230)
• Masks have a quicklook method (#228)
• Memory mapping can be disabled (#226)
• Add xor operations for Masks (#241)
• Added a new StokesSpectralCube class to deal with 4-d cubes (#249)

0.2.2 (2015-03-12)

• Output mask as a CASA image #171
• ytcube exports to .obj and .ply too #173
• Fix air wavelengths, which were mistreated (#186)
• Add support for sum/mean/std over both spatial axes to return a OneDSpectrum object. This PR also removes numpy 1.5-1.7 tests, since many spectral_cube functions are not compatible with these versions of numpy (#188)

0.2.1 (2014-12-03)

• CASA cube readers now compatible with ALMA .image files (tested on Cycle 2 data) #165
• Spectral quicklooks available #164 now that 1D slices are possible #157
• to_pvextractor tool allows easy export to pvextractor #160
• to_glue sends the cube to glue #153

0.2 (2014-09-11)

• moments preserve spectral units now #118
• Initial support added for Air Wavelength. This is only 1-way support, round-tripping (vacuum->air) is not supported yet. #117
• Integer slices (single frames) are supported #113
• Bugfix: BUNIT capitalized #112
• Masks can be any array that is broadcastable to the cube shape #115
• Added .header and .hdu convenience methods #120
• Added public functions apply_function and apply_numpy_function that allow functions to be run on cubes while preserving important metadata (e.g., WCS)
• Added a quicklook tool using aplpy to view slices (#131)
• Added subcube and ds9 region extraction tools (#128)
• Added a to_yt function for easily converting between SpectralCube and yt datacube/dataset objects (#90, #129)
• Masks' .include() method works without data arguments. (#147)
• Allow movie name to be specified in yt movie creation (#145)
• add flattened_world method to get the world coordinates corresponding to each pixel in the flattened array (#146)

0.1 (2014-06-01)

• Initial Release.