Skip to content

Releases: jorenham/Lmo

v0.11.1

05 Nov 23:40
43b68c5
Compare
Choose a tag to compare

Bugfix release

What's Changed

  • Ruff 0.1.4 upgrade and config finetuning by @jorenham in #68
  • Prevent IntegerOverflow on windows: prefer np.int64 over np.int_ by @jorenham in #69
  • Consistent tie-breaking in concomitants with lexical ordering by @jorenham in #72
  • Fix incorrect trimming of extreme values, including inf by @jorenham in #73
  • Add missing @overloads for L-moment aliases by @jorenham in #75
  • Added type annotations for passing (symmetric) trim as scalar by @jorenham in #77

Full Changelog: v0.11.0...v0.11.1

v0.11.0

30 Oct 03:09
41f0143
Compare
Choose a tag to compare

What's Changed

Full Changelog: v0.10.1...v0.11.0

v0.10.2

02 Oct 20:55
d6c2ded
Compare
Choose a tag to compare

What's Changed

Full Changelog: v0.10.1...v0.10.2

v0.10.1

11 Sep 21:49
997da03
Compare
Choose a tag to compare

What's Changed

New Contributors

Full Changelog: v0.10.0...v0.10.1

v0.10.0

11 Sep 01:26
0003fe1
Compare
Choose a tag to compare

What's Changed

  • Added more linting, fixed many code style issues and a few typos by @wolph in #13
  • L-moment covariance matrix from CDF by @jorenham in #14
  • L-stats covariance matrix from CDF by @jorenham in #15
  • Calculate theoretical L-* from scipy.stats.rv_continuous distributions by @jorenham in #16
  • Python 3.12 support by @jorenham in #17
  • Tighter bounds on the (absolute) L-moment and L-ratio by @jorenham in #18
  • Improved L-stats API by @jorenham in #19
  • Goodness-of-fit testing of any distribution, using L-moments or L-stats by @jorenham in #20
  • Standardized optional scipy.integrate.quad options in lmo.theoretical by @jorenham in #21
  • Theoretical influence functions & robustness measures by @jorenham in #22
  • lmo.theoretical: Discrete distributions, numerical accuracy improvements, and cleaner code by @jorenham in #25
  • Add theoretical L-* methods to all scipy.stats univariate distributions by @jorenham in #26
  • Method of L-moments (LMM) by @jorenham in #27
  • Empirical influence functions by @jorenham in #28
  • Allow passing trim=t as alias for trim=(t, t) by @jorenham in #29

New Contributors

  • @wolph made their first contribution in #13

Full Changelog: v0.9.0...v0.10.0

v0.9.0

09 Aug 01:57
9ef91e5
Compare
Choose a tag to compare

What's Changed

Full Changelog: v0.8.0...v0.9.0

v0.8.0

24 Jul 02:25
86cb536
Compare
Choose a tag to compare
  • A novel generalization of trimmed L-moments and L-comoments: Fractional trimming 🎉 . Trim lengths can now be any positive float. (#11)
  • New convenience methods: lmo.l_stats, lmo.l_costats, and lmo.theoretical.l_(ratio|stats)_from_(cdf|ppf)
  • New low-level lmo.ostats module, for internally used (fractional) order-statistics calculations.
  • Improved function typing with typing.overload .
  • Improved parameter typing by annotating **kwargs with TypedDict.
  • Many documentation improvements.
  • Reduced test flakyness, removed redundant tests, and added some new ones.
  • Made the internal low-level _pwm module public, by renaming it to pwm_beta.
  • BREAKING: most dtype and axis kwargs are now keyword-only arguments.

Full Changelog: v0.7.0...v0.8.0

v0.7.0

17 Jul 21:40
15d913b
Compare
Choose a tag to compare
  • Improved README.md [#1, #2]
  • Theoretical L-moments from univariate continuous distributions, using either a CDF of PPF [#6]
  • Improved numerical stability of sample L-(co)moments with several orders of magnitude
  • General documentation improvements
  • Improved code style, by enforcing a stricter ruff ruleset
  • [BREAKING] Bumped minimum numpy version to 1.22
  • [BREAKING] Bumped minimum scipy version to 1.9

Full Changelog: v0.6.1...v0.7.0

v0.6.1

03 Jul 13:40
f150b4a
Compare
Choose a tag to compare

Bugfix release: fix cache=True if r is larger than what's cached

Full Changelog: v0.6.0...v0.6.1

v0.6.0

03 Jul 01:36
0113cf7
Compare
Choose a tag to compare
  • Optional cache=True kwarg to speed up consecutive L-moment calculations

Full Changelog: v0.5.3...v0.6.0