1.4.1

2026-03-15T01:24:07Z

Bug fixes:

Fix test_hexadecimal_with_libc_bulk(), see #1049 (Doug Torrance)
Keep available deprecated aliases for mpc/mpf_log() (Sergey B Kirpichev)
Use version_file option of setuptools-scm to keep version info, see #1048
(Sergey B Kirpichev)
Add workaround for test on s390x, see #1061 (Sergey B Kirpichev)

Full Changelog: 1.4.0...1.4.1

1.4.0

2026-02-23T05:35:32Z

Features:

Support underscores as digit separators per PEP 515, see #661 (Sergey B Kirpichev)
Add rationals converter for mpf's, see #666 (Sergey B Kirpichev)
Rewrite bernpoly/eulerpoly to avoid dependency on bernoulli(1) convention, see #700 (Sergey B Kirpichev)
Support base kwarg for from_str(), see #703 (Jonathan Warner, Sergey B Kirpichev)
Support randmatrix() for mp.iv and mp contexts, see #527 (Maximilian Gaukler)
Added rank() function for matrices, see #610 (Jan-Philipp Hoffmann)
Add plus flag to select the B_1 sign convention for bernoulli/bernfrac, see #724 (Jeremy Tan Jie Rui, Sergey B Kirpichev)
Add mpf.as_integer_ratio() method, support construction of mpf from Decimal objects, see #731 (Sergey B Kirpichev)
Expose lower/upper_gamma functions, see #740 (Sergey B Kirpichev)
Support mpc initialization from string, see #743 (Sergey B Kirpichev)
Support asinh/acosh/atanh in the fp context, see #750 (Sergey B Kirpichev)
Support binary/octal/hexadecimal string output, see #711 (Jonathan Warner, Sergey B Kirpichev)
Support pickling for matrices and mpi, see #761 (Sergey B Kirpichev)
Support matrix.__array__() dunder method, see #767 (Sergey B Kirpichev)
Support more number syntaxes, see #778 (Sergey B Kirpichev)
Run mpmath as a module for interactive work, see #773, #923, #931, #936, #939 and #954 (Sergey B Kirpichev)
Add signed option to to_man_exp(), see #783 (Sergey B Kirpichev)
Add fp.hypot, see #798 (Sergey B Kirpichev)
Support inf/nan's in ctx.almosteq(), #802 (Sergey B Kirpichev)
Implement mpf.__format__(), see #819, #831, #850, #859, #857, #862, #881, #944 and #966 (Javier Garcia, Sergey B Kirpichev)
Support conversion from scalar ndarray's, see #821 (Sergey B Kirpichev)
Support rounding modes in mpf.__format__, see #823, #831, #834 and #969 (Javier Garcia, Sergey B Kirpichev)
Support '%' presentation type for mpf, see #847 (Sergey B Kirpichev)
Support gmpy2-like rounding modes in to_str(), see #830 (Javier Garcia)
Implement 'a'/'A' formating types for mpf.__format__, see #841 and #870 (Sergey B Kirpichev)
Add mpc.__format__(), see #855 (Sergey B Kirpichev)
Now mpf.__round__() returns mpf, see #826 and #966 (Sergey B Kirpichev)
Support 'b' (binary) format type for mpf/mpc, see #867 (Sergey B Kirpichev)
Implement mpf.__floordiv__() and mpf.__divmod__(), see #873 (Sergey B Kirpichev)
Add parameters for MPContext constructor, see #876 and #963 (Sergey B Kirpichev)
Add MPFR-compatible aliases for rounding modes, see #892 (Sergey B Kirpichev)
Support negative indexes in matrix, see #897 (Riccardo Orsi)
Better introspection support for decorated functions, see #900 (Sergey B Kirpichev)
Add moving sofa demo, see #924 (Sergey B Kirpichev)
Support spherical Bessel functions (jn/yn), #935 (Sergey B Kirpichev)
Add pretty_dps context property to control number of printed digits, see #933 (Sergey B Kirpichev)
Support thousands separators for formatting of fractional part, see #925 and #936 (Sergey B Kirpichev)
Use PyREPL, as fallback (no IPython), see #941 (Sergey B Kirpichev)
Add exp2() and log2(), see #948 (Sergey B Kirpichev)
Support rounding property for the mp context, see #963 (Sergey B Kirpichev)
Add Fox H-function with rational A/B parameters (foxh()), see #982 (Hongren Zheng)
Provide experimental support for free-threading builds, see #993 (Sergey B Kirpichev)

Compatibility:

Drop Python 2 support, see #629 (Fangchen Li)
Drop support for Python versions < 3.9, see #675 and #911 (Sergey B Kirpichev)
Drop private mpq class, use Rational's, provided by backend, see #691 and #769 (Sergey B Kirpichev)
Drop to_pickable()/from_pickable() helpers, see #667 and #769 (Sergey B Kirpichev)
Direct access to _mpf_ tuples now deprecated, please use from/to_man_exp() functions and special constants (finf, fninf and fnan), see #783 (Sergey B Kirpichev)
Removed sage backend, see #732 (Sergey B Kirpichev)
Drop MPMATH_STRICT environment variable, see #759 (Sergey B Kirpichev)
Deprecate current (descending) order of coefficients in polyval(), etc, see #779, #844 and #845 (Sergey B Kirpichev, Warren Weckesser)
Deprecate mpmath.math2, see #769 (Sergey B Kirpichev)
Importing from the mpmath.libmp submodules is deprecated, use instead from mpmath.libmp import foo, see issue #704 (comment) for available functions (Sergey B Kirpichev)
Deprecate bitcount function, see #721 and #955 (Sergey B Kirpichev)
Deprecate mpf/mpc_log, see #989 (Sergey B Kirpichev)
Deprecate fp.is_special(), see #1042 (Sergey B Kirpichev)

Bug fixes:

sum_accurately(), betainc() and power() fixes, see #664 (Sergey B Kirpichev)
Warn users about Python's true division, see #670 (Sergey B Kirpichev)
Propagate nan's in ei_asympt(), see #672 (Sergey B Kirpichev)
Fix matrix.__eq__, fix string parsing with underscores, see #679 (Sergey B Kirpichev)
Raise IndexError if matrix index out of bounds, see #689 (Sergey B Kirpichev)
Fix nan handling in fp.mag() and hyper(), see #688 (Sergey B Kirpichev)
Optimize sparse matrix dot product, see #450 (Tyler Chen)
Correct pow() for mpf's to be consistent with mpfr/float's, see #690 (Sergey B Kirpichev)
Improve hypsum non-convergence behaviour, see #703 (Benjamin Fischer)
Fixed TypeError in LU_comp, see #610 (Jan-Philipp Hoffmann)
Fix disagreement fp.mod vs mp.mod, see #710 (Sergey B Kirpichev)
Skip eigenvectors if left=right=False for one-dimentional matrix, see #713 (Sergey B Kirpichev)
Fix mpc() constructor to be compatible with complex(), fix disagreement fp.pow vs mp.pow, see #731 (Sergey B Kirpichev)
Add derivative keyword argument for besselk(), see #735 (Sergey B Kirpichev)
Fix quadosc(), fast case in _hyp2f1(), correct fp.gamma() and fp.isnprint(), see #740 (Sergey B Kirpichev)
Pevent erroneous setting of dps/prec on mpmath module, see #678 (Colin B. Macdonald)
Update splot() for recent matplotlib, see #747 (Sergey B Kirpichev)
acos_asin(): don't try to normalize special numbers, fix repr(mp.eps), see #750 (Sergey B Kirpichev)
Fix choleksy_solve for complex matrix, see #755 (Qiming Sun)
Fix hang in polylog_general(), fix demos, use cholesky_solve() for overdetermined complex linear systems, fix several issues with empty matrices, see #759
sqrt(z): special case for infinite z.imag, see #777 (Sergey B Kirpichev)
Correct atan2(±inf, ±inf), see #775 (Sergey B Kirpichev)
Handle infinite arguments in tan/tanh, see #785 (Sergey B Kirpichev)
Drop mpf.__complex__(), workaround 1/z division for acot/asec/acsc/acoth, see #797 (Sergey B Kirpichev)
Handle more special cases in besselk and hyp1f1, see #792, #800 and #801 (Sergey B Kirpichev)
Correct asin/acos for infinite arguments, see #795 (Sergey B Kirpichev)
Reduce memory usage for QuadratureRule cache, see #812 (Sergey B Kirpichev)
Drop sign from nstr(mpf('inf')) output, see #828 (Sergey B Kirpichev)
Improve accuracy of log1p(), see #803 and #854 (Tim Peters, Sergey B Kirpichev)
Normalize mpf in mp.npconvert(), fix special cases in bernpoly(), see #839 (Sergey B Kirpichev)
Fix mpf_div() for prec=0, see #849 (Sergey B Kirpichev)
Raise ValueError for complex infininity condition in zeta(s, a), see #864 (Sergey B Kirpichev)
Enable trap_complex for MDNewton, see #870 (Sergey B Kirpichev)
Use correct mixed-mode functions in fsub/fdiv, add special cases for infinities in mpc_*div(), see #873 (Sergey B Kirpichev)
Revert "fix ellippi to return an inf instead of raising an exception", see #875 (Sergey B Kirpichev)
Reject invalid strings in from_str(), see #886 (Sergey B Kirpichev)
Use parity formula for besseli, see #889 (Sergey B Kirpichev)
Special case in ctx.hypsum for infinite z, see #902 (Sergey B Kirpichev)
Raise an exception if iv's comparison can't be decided, see #903 (Sergey B Kirpichev)
Add special case for ±inf in polylog_continuation(), see #904, #1034 and #1037 (Sergey B Kirpichev, Colin B. Macdonald)
Increase working precision in polylog_general() for negative s, see #898 (Sergey B Kirpichev)
Correct case for integer n in besselj/besseli, see #909 (Sergey B Kirpichev)
Use sum_accurately() in hankel1/2(), see #926 (Sergey B Kirpichev)
Ensure mpf_bernoulli() returns normalized answer, see #939 (Sergey B Kirpichev)
Use mpf_log1p in acos_asin() helper (implementing Hull et al algorithm), see #948 and #1036 (Sergey B Kirpichev)
Fix kwargs passing in the nstr() for mpc, see #964 (David Walker)
Fix exception type for int(inf), see #966 (Sergey B Kirpichev)
Ensure exp, sin, tan, etc have a correct __name__ attribute, see #997 (Warren Weckesser)
Matrix raise ValueError in case of negative dimensions, see #1004 (Ayush Baranwal)
Support lists in sinm() and cosm(), see #1003 (Ayush Baranwal)
Properly handle nan's elliprj(), see #1038 (Sergey B Kirpichev)
Raise ValueError for logm(0), see #1017 (Ayush Baranwal)
Fix erf(z) with re(z) of large magnitude, see #1039 (Sergey B Kirpichev)
Return nan's for polylog(s, nan) or polylog(s, nan+nanj), see #1041 (Sergey B Kirpichev)
Fix fp.isnormal() for subnormals, see #1042 (Sergey B Kirpichev)

Maintenance:

Use codecov/coverage-action, see #674 (Sergey B Kirpichev)
Run CI tests with pytest-xdist, see #685 (Sergey B Kirpichev)
Add pyproject.toml, depend on flake518, see #684 (Sergey B Kirpichev)
Port torture.py/extratest_zeta.py/extratest_gamma.py to the pytest framework, see #687 (Sergey B Kirpichev)
Create CITATION.bib, see #681 (Devharsh Trivedi)
Avoid using star imports in tests and documentation, see #698 (Sergey B Kirpichev)
Fix some py2's remnants and remove old gmpy workarounds, see #699 (Sergey B Kirpichev)
Use math.isqrt in isqrt_python calculations, see #695 (Daiki Takahashi)
Use gcd() and other bigint's functions from the backend, see #697 (Sergey B Kirpichev)
Drop legacy and redundant code, see #701 (Sergey B Kirpichev)
Change FPContext to use more functions from the stdlib, see #692 (Sergey B Kirpichev)
Avoid dynamic method creation in _mpf, see #702 (Sergey B Kirpichev)
Enable testing on CPython 3.12-dev, see #706 (Sergey B Kirpichev)
Use bit_length() method instead of bitcount(), see #721 (Sergey B Kirpichev)
Use lru_cache() in ifib() and eulernum(), isprime() alternatives from backends, see #722 (Sergey B Kirpichev)
Use setuptools_scm to update __version__, see #694 (Sergey B Kirpichev)
Run tests on 3.13, see #759 (Sergey B Kirpichev)
Do not build depend on pip and wheel, see #758 (Gonzalo Tornaría)
Add CONTRIBUTING.rst, see #763 (Sergey B Kirpichev)
Simplify ctx_mp_python.py, see #806 (Sergey B Kirpichev)
Update gmpy2 deps, see #808 and #813 (Sergey B Kirpichev)
Enable testing on 3.14, see #851 (Sergey B Kirpichev)
Refactor Github Actions, see #905 (Sergey B Kirpichev)
Build and publish wheel, see #913 (David Hotham)
Use the intended setuptools_scm integration pattern, see #940 (Ronny Pfannschmidt)
Add backport action, see #1042 (Sergey B Kirpichev)

See the release milestone (1.4) for a complete list of issues and pull requests involved in this release.

Full Changelog: 1.3.0...1.4.0

1.4.0b4

2026-02-16T06:59:16Z

1.4.0b4

1.4.0b3

2026-01-03T01:47:39Z

1.4.0b3

1.4.0b2

2025-10-13T05:58:35Z

1.4.0b2

1.4.0b1

2025-09-06T02:47:49Z

1.4.0b1

1.4.0a8

2025-09-01T04:44:55Z

1.4.0a8

1.4.0a7

2025-08-08T08:23:34Z

1.4.0a7

1.4.0a6

2025-07-03T12:59:58Z

1.4.0a6

1.4.0a5: Merge pull request #951 from skirpichev/gmp-backend

2025-04-30T07:33:56Z

Add python-gmp as optional backend (gmpy2 alternative)

Release notes from mpmath

1.4.1

1.4.0

1.4.0b4

1.4.0b3

1.4.0b2

1.4.0b1

1.4.0a8

1.4.0a7

1.4.0a6

1.4.0a5: Merge pull request #951 from skirpichev/gmp-backend