What's Changed

• Remove some duplicated code in photon phasoegrams. by @kerrm in #1904
• Don't subtract mean by default while simulating TOAs by @abhisrkckl in #1909
• Type hints in pint.fitter by @abhisrkckl in #1885
• Power-law solar wind GP as a noise component by @jeremy-baier in #1854
• Explicitly specify Tspan in PL*Noise components by @abhisrkckl in #1908
• Fixed nan usage by @dlakaplan in #1915
• Make tests run in parallel by @abhisrkckl in #1913
• Fix wideband DM noise simulations + Utility functions for constructing wideband residuals, scaled uncertainties, and design matrix by @abhisrkckl in #1910
• find_empty_masks can now handle models with SWX by @dlakaplan in #1924
• Fixes binary conversion errors by @dlakaplan in #1922
• Update utils.taylor_horner_deriv documentation example to give the co… by @Jackson-D-Taylor in #1919
• Add TNCHROMIDX directly while using cmwavex_setup by @abhisrkckl in #1918
• Calculate integer phases when using polycos in photonphase by @matteobachetti in #1930
• noise_resids in WidebandTOAResiduals by @abhisrkckl in #1932
• Chromatic exponential dip by @abhisrkckl in #1925
• Remove broken script event_optimize_multiple by @abhisrkckl in #1933
• Remove ORBWAVE parameters from BinaryBTPiecewise by @abhisrkckl in #1937
• Fix pintk.Pulsar.print_chi2() by @abhisrkckl in #1939
• Fix tcb2tdb factors for solar wind GP parameters by @abhisrkckl in #1943
• Include DM information in TOAs.get_summary() for wideband TOAs by @abhisrkckl in #1940
• Make whitened residuals available pre-fit + Whitened residuals for wideband by @abhisrkckl in #1935

New Contributors

• @jeremy-baier made their first contribution in #1854

Full Changelog: 1.1.3...1.1.4

[1.1.3] 2025-04-25

Changed

• In Residuals, store correlated noise amplitudes instead of noise residuals. Residuals.noise_resids is now a @property.
• Refactor pint.fitter to reduce code duplication

Added

• Simulate correlated DM noise for wideband TOAs
• Type hints in pint.models.timing_model
• full_designmatrix() and full_basis_weights() methods in TimingModel
• Added checkbox for optional subtraction of mean in pintk
• Added Log-Linear Powerlaw noise parameters to PLRedNoise, PLDMNoise, PLChromNoise
• TimingModel.ntmpar property
• Location of NENUFAR

Fixed

• Shape of Fitter.resids.noise_ampls (it was wrong before due to bad indexing)
• Made TimingModel.is_binary() more robust.
• Correct value of (1/year) in powerlaw() function
• Fixed TestPintk
• Fixed the noise realization indexing in Fitters
• Added missing newline in tempo_polyco_table_writer()
• Proper warning in read_polyco_file()

Removed

• Definition of @cached_property to support Python<=3.7
• The broken data.nanograv.org URL from the list of solar system ephemeris mirrors
• Broken fitter class CompositeMCMCFitter (this fitter was added seemingly to deal with combined radio and high-energy datasets, but has since been broken for a while.)

Full Changelog: 1.1.1...1.1.3

What's Changed

• Big change is bumping oldest Python supported to 3.9 and bumping minimum versions of numpy to 1.23.0, astropy to 5.0.5, scipy to 1.9.0, and matplotlib to 3.4.3

Full Changelog: 1.0.1...1.1

What's Changed

• Bayesian information criterion by @abhisrkckl in #1741
• Allow "CLK UNCORR" in par files by @abhisrkckl in #1747
• Make Astrometry & SolarSystemShapiro faster by @abhisrkckl in #1748
• Fix split swx by @LuluAgazie in #1753
• Allow pintk to read par files with UNITS TCB and BINARY T2 + Exclude function parameters in pintk by @abhisrkckl in #1750
• Fix #1759 (coordinate comparison identity) by @dlakaplan in #1761
• Updated location of CCERA by @dlakaplan in #1758
• Custom type definitions by @dlakaplan in #1756
• New DMX setup function and type hinting for almost all of utils.py by @dlakaplan in #1752
• Try citation.cff by @dlakaplan in #1767
• Fixed bug in residual calculation when adding or removing phase wraps… by @scottransom in #1732
• Better algorithm for TCB -> TDB conversion by @abhisrkckl in #1765
• Type hints in pint.toa and get_model & get_model_and_toas functions by @abhisrkckl in #1755
• Moved include_gps and include_bipm out of Observatory objects by @paulray in #1763
• Whitened residuals in pintk by @abhisrkckl in #1775
• Chromatic delay similar to dispersion delay by @abhisrkckl in #1616
• Remove debugging statement by @dlakaplan in #1778
• Minimal fixes for numpy 2.0 by @dlakaplan in #1784
• Updates for scipy 1.14 by @dlakaplan in #1793
• Fixed docstring in one simulation routine by @dlakaplan in #1782
• TOAs.get_Tspan() method by @abhisrkckl in #1781
• Fix removal of top-level parameter by @dlakaplan in #1794
• Handle BIPM and wideband DMs correctly in make_fake_toas_fromtim by @abhisrkckl in #1790

Full Changelog: 1.0...1.0.1

What's Changed

• Fixed README for runtime data by @dlakaplan in #1685
• Added convert_parfile to list of command-line tools in RTD by @dlakaplan in #1687
• Fix MCMC_walkthrough.ipynb by @dlakaplan in #1679
• Fixed bug in derived params when OMDOT uncertainty is 0 by @dlakaplan in #1689
• Fix bug in make_fake_toas_fromtim by @dlakaplan in #1698
• Fix observatory.init.py by @dlakaplan in #1696
• Plot model DM in pintk + more tests for pintk by @abhisrkckl in #1678
• Moved get_derived_params to timing_model by @dlakaplan in #1692
• DDH Binary model by @dlakaplan in #1693
• Maximum-likelihood fit for ECORR by @abhisrkckl in #1673
• Consistent naming in get_params_mapping by @abhisrkckl in #1704
• Better exceptions for unimplemented binary models (BTX etc) by @abhisrkckl in #1701
• CI ephemeris test no longer requires access to static NANOGrav site by @dlakaplan in #1706
• change_binary_epoch in TimingModel.compare() by @abhisrkckl in #1703
• Component validation for WaveX and DMWaveX by @abhisrkckl in #1702
• get_observatory shouldn't overwrite include_bipm and include_gps unless given explicitly by @abhisrkckl in #1711
• Convert WaveX to PLRedNoise and DMWaveX to PLDMNoise by @abhisrkckl in #1694
• Added spacecraft as alias for geocenter by @dlakaplan in #1724
• Allow simulated TOAs to maintain a non-zero mean by @dlakaplan in #1717
• Improve clock error reporting by @aarchiba in #1720
• Allow very old Princeton format TOAs by @scottransom in #1722
• Remove ERFA warnings by @aarchiba in #1721
• Removed np.compat.long (deprecated) by @dlakaplan in #1728
• Fix #1729 by @dlakaplan in #1730
• Changed error bars to scaled for pintk by @dlakaplan in #1735
• Fixed common CI failure in tests/test_observatory.py::test_json_observatory_input_latlon by @dlakaplan in #1738
• Updated docs for ELL1H by @dlakaplan in #1736
• FDJUMPDM : System-dependent DM offset by @abhisrkckl in #1731
• Proper motion conversion/calculations can now uniformly use float, Quantity, or Time by @dlakaplan in #1737
• Try macos 12 to see if it is still intel by @dlakaplan in #1745
• Guess binary model & add conversion script by @vhaasteren in #1695

Full Changelog: 0.9.8...1.0

What's Changed

• Fix sphinx build by specifying theme explicitly by @dlakaplan in #1627
• Setting model.PARAM1 = model.PARAM2 no longer overrides the name of PARAM1 by @dlakaplan in #1624
• Fix .value() for pairParameters by @dlakaplan in #1628
• Remove an undefined name by @munahaf in #1631
• Remove an undefined name by @munahaf in #1630
• Replace a test expression to remove a logical short circuit by @munahaf in #1629
• Correction to docstring for pbprime functions by @coclar in #1606
• Fix conversion from ICRS -> ECL when pm uncertainties are None by @dlakaplan in #1632
• Remove GLSFitter dependence from Residuals by @abhisrkckl in #1618
• Addition of WaveX model by @LuluAgazie in #1609
• Mixed wideband + narrowband TOAs objects should not be allowed. by @abhisrkckl in #1636
• Radial velocities by @dlakaplan in #1638
• Publication output by @abhisrkckl in #1621
• DMWaveX model for Fourier representation of DM noise by @abhisrkckl in #1637
• Include CHI2 in par output by @dlakaplan in #1634
• Bayesian interface for Wideband data (No correlated noise) by @abhisrkckl in #1426
• Updated piecewise orbital model by @poneill129 in #1545
• Simulate correlated noise by @abhisrkckl in #1645
• Fix versioneer by @paulray in #1650
• Sensibly handle unfittable params, No unfittable params in plk by @abhisrkckl in #1648
• Fixed units on errors in fake TOAs by @dlakaplan in #1653
• Update to python 3.12 for testing by @dlakaplan in #1652
• Faster implementations of astrometric ssb calculations by @dlakaplan in #1646
• Fit white noise parameters by @abhisrkckl in #1639
• Added elongation for pintk by @dlakaplan in #1655
• Speed up some tests by @dlakaplan in #1657
• pintk can swap inputs by @dlakaplan in #1656
• Disabled axes for pintk when not in timing model or data by @dlakaplan in #1663
• Add full name for observatories by @dlakaplan in #1668
• Fix sign of y coordinate for Pico Veleta so it's no longer in the sea by @dlakaplan in #1672
• Make whitened residuals available by @abhisrkckl in #1671
• Docstring for TimingModel.get_deriv_funcs by @abhisrkckl in #1675
• Fix examples by @abhisrkckl in #1674
• More realistic simulation of narrowband TOAs of multiple frequencies by @abhisrkckl in #1670
• Plot wideband DMs and DM residuals in pintk by @abhisrkckl in #1534
• Simulate TOA flags by @abhisrkckl in #1682
• Set UNITS to TDB if it is not given in the par file. by @abhisrkckl in #1681
• Add flags example notebook by @dlakaplan in #1683

New Contributors

• @munahaf made their first contribution in #1631
• @coclar made their first contribution in #1606
• @LuluAgazie made their first contribution in #1609

Full Changelog: 0.9.7...0.9.8

What's Changed

• Implemented Roemer delay terms up to O(e^3) in ELL1 model by @wcfiore in #1573
• Deleting JUMP1 does not break fitting by @dlakaplan in #1604
• Make the usage of TZR TOA explicit by @abhisrkckl in #1607
• New pint.print_info() function, Update CONTRIBUTING.rst by @abhisrkckl in #1601
• Simulating TOAs from tim file when PLANET_SHAPIRO is true now works by @dlakaplan in #1614
• Autocorrelation check for event_optimize by @devbhakt in #1610
• Download data file automatically in high_level_benchmark.py by @abhisrkckl in #1619
• Make TimingModel.params and TimingModel.ordered_params identical. by @abhisrkckl in #1617
• Minor doc updates by @dlakaplan in #1620
• FDJUMP implementation (Hacky version) by @abhisrkckl in #1615

New Contributors

• @wcfiore made their first contribution in #1573

Full Changelog: 0.9.6...0.9.7

What's Changed

• Sourcery refactors III: Residuals, Fitters, Sampler, random models, events etc. by @abhisrkckl in #1563
• Sourcery refactors IV: Scripts by @abhisrkckl in #1564
• Sourcery refactors V: Observatories, Models by @abhisrkckl in #1565
• Validate TimingModel components by @abhisrkckl in #1553
• Fixed docstring in add_param_from_top by @dlakaplan in #1572
• Roundtrip test to make sure clock corrections are not written to tim file by @dlakaplan in #1571
• Fix logging when using gridding by @dlakaplan in #1435
• Explicit phase offset as a fittable parameter by @abhisrkckl in #1557
• Tool to get units associated with any parameter by @dlakaplan in #1574
• Fix test_model_derivatives.py failure by @dlakaplan in #1584
• Move unreleased changelog to a separate file by @abhisrkckl in #1575
• Add import of matplotlib figure by @matteobachetti in #1579
• Include list of observatories in docs by @dlakaplan in #1581
• Override/add to par file when building models by @dlakaplan in #1568
• Updated docs for Apple silicon by @dlakaplan in #1588
• Update README by @paulray in #1590
• Changed default toa errors for event files. can also set explicitly by @dlakaplan in #1591
• Allow merging a list of TOAs objects with only one element by @rossjjennings in #1593
• Standardize tests and fix some deprecation warnings by @paulray in #1595
• Sampler.chain deprecated fix by @devbhakt in #1596
• Model conversion ICRS <-> ECL works if PM uncertainties are not set by @dlakaplan in #1598

Full Changelog: 0.9.5...0.9.6

[0.9.5] 2023-05-01

Changed

• Changed minimum supported version of scipy to 1.4.1
• Moved DMconst from pint.models.dispersion_model to pint to avoid circular imports
• Removed references to astropy._erfa (removed since astropy 4.2)
• Refactor Dre method, fix expressions for Einstein delay and post-Keplerian parameters in DD model
• Updated contributor list (AUTHORS.rst)
• Emit an informative warning for "MODE" statement in TOA file; Ignore "MODE 1" silently
• Version of sphinx-rtd-theme updated in requirements_dev.txt
• Updated black version to 23.x
• Older event loading functions now use newer functions to create TOAs and then convert to list of TOA objects
• Limited hypothesis to <= 6.72.0 to avoid numpy problems in oldestdeps

Added

• Documentation: Explanation for DM
• Methods to compute dispersion slope and to convert DM using the CODATA value of DMconst
• TimingModel.total_dispersion_slope method
• Explicit discussion of DT92 convention to DDK model
• HAWC, HESS and ORT telescopes to the list of known observatories
• Documentation: making TOAs from array of times added to HowTo
• Method to make TOAs from an array of times
• Clock correction for LEAP
• Wideband TOA simulation feature in pint.simulation and zima
• ELL1k timing model
• Test for MCMCFitter
• Added multiprocessing capability to event_optimize
• Can merge TOAs using '+' operator; in-place merge using '+=' operator
• funcParameters defined as functions operating on other parameters
• Option to save emcee backend chains in event_optimize
• Documentation on how to extract a covariance matrix
• DDS and DDGR models
• Second-order corrections included in ELL1
• Module for converting between binary models also included in convert_parfile
• Method to get a parameter as a uncertainties.ufloat for doing math
• Method to get current binary period and uncertainty at a given time regardless of binary model
• TCB to TDB conversion on read, and conversion script (tcb2tdb)
• Functions to get TOAs objects from satellite data (Fermi and otherwise)
• Methods to convert a TOAs object into a list of TOA objects

Fixed

• Syntax error in README.rst
• Broken notebooks CI test
• BIPM correction for simulated TOAs
• Added try/except to test_pldmnoise.py/test_PLRedNoise_recovery to avoid exceptions during CI
• Import for longdouble2str in get_tempo_result
• Plotting orbital phase in pintk when FB0 is used instead of PB
• Selection of BIPM for random models
• Added 1 sigma errors to update the postfit parfile errors in event_optimize
• Fixed DDS CI testing failures
• Add SolarSystemShapiro to the timing model only if an Astrometry component is present.

Full Changelog: 0.9.3...0.9.5