* a few floating-point error shims to allow our
testsuite to pass for MacOS arm64 with the recen...

MAINT: signal: replace `np.r_["-1", arrays]` by a more sane incantation

MAINT:ENH:sparse.linalg: Rewrite iterative solvers in Python, remove FORTRAN code

* Fixes #18691

* as noted there, GPG signing of our PyPI binaries
and sdists is pointless and j...

Provide several hooks which a subclass can override. The use case is repeated evaluation of an i...

Apparently, if the first arg of np.r_ is a string, it's the axis. Ugh.
So remove that.

Also, al...

* DOC: stats.genlogistic: note Type I generalized logistic / skew logistic

DEV: Add ability to run memory benchmarks on macOS

[skip ci]

Before, a scalar value for sigma raised a ValueError:

```python
>>> curve_fit(..., sigma=2.0)
V...

Add new method of finding peak memory usage using `os.wait4()`.
This also gives the true peak us...

Closes #18452

Fixes orthonormalization in DSTs type 2 and 3

More recent versions cause deprecation warnings to be emitted from
`pkg_resources` by (at least...

* CI: remove setuptools install from musllinux job

No longer used/needed, since we removed th...

* Fixes #18595

* use `git apply` to directly apply the patch from
gh-18595 that worked durin...

[skip ci]

---------

Co-authored-by: Ralf Gommers <[email protected]>

* ENH: stats.dirichlet: add covariance method

[skip ci]

MAINT: stats.truncnorm/stats.betaprime: fix _munp for higher moments, add tests

* ENH: Adding ?tgsyl_lwork routine templates for real types

* ENH: Fixed the _lwork subroutin...

[skip ci]

Co-authored-by: Pamphile Roy <[email protected]>

TST: fix two test failures that showed up on conda-forge

The ppc64le one is a small test bump. The PyPy-on-Windows one
is a real bug in PyPy, but it's no...

* ENH:optimize:Convert nnls solver code to Python

* MAINT:optimize:Remove replaced nnls FORTR...

* BUG: Fixes how the length of argument x of linalg.blas.?gbmv is checked

Given an m×n banded...

* ENH: stats.beta.entropy: asymptotic expansion to avoid precision loss for large a and b

* ENH: stats.multivariate_t: add asymptotic expansion to avoid precision loss

Add NdBSpline: n-dim tensor product b-spline object

* ENH: loggamma.entropy: override for improved accuracy

Co-authored-by: Daniel Schmitz <4065...

ENH: stats.CovViaCholesky: make `_covariance` a cached property

[skip ci]

Co-authored-by: Jake Bowhay <[email protected]>

Co-authored-by: Philippe DONNAT <[email protected]>

[skip actions] [skip circle]

Mark `void` functions as `noexcept` in `_rotation.pyx`

Co-authored-by: Atsushi Sakai <[email protected]>

DOC: stats.loguniform._sf: add comment about not overriding

ENH: stats.loglaplace.sf: override for accuracy in right tail

MAINT: interpolate: delete duplicated FITPACK bisplev interface

* improve an error message based on code review

[skip ci]

Co-authored-by: Evgeni Burovski <...

* TST: Enable LSODA tests and add failing test.

* ENH: Select appropriate segment for each me...

* DOC: more explicit example usage of scipy.linalg.lu_factor

* Document the 1-based to 0-base...

* ENH: stats.pearson3: override for improved accuracy in right tail
* TST: stats.pearson3: sim...

ENH: Faster _select_samples in _differential_evolution.py

* ENH: stats.rdist.sf: override for accuracy in right tail

* BUG: Fix fails of some tests and bad behaviour for x=zeta in levy_stable

* Reset parameters...

[skip ci]

* DEP: remove parameter centered of stats.qmc.LatinHypercube

* DOC: remove reference to cente...

REL: set version to 1.12.0.dev0

DOC: SciPy 1.11.0 release notes

* formatting cleanup for `sparse` section, and
move deprecation information to the appropriate
s...

* address suggestions from code review

[skip actions] [skip cirrus]

* apply reviewer suggestions in cases where it was possible to do
directly via the GH interfac...

* don't merge until the `1.11.0` release notes have
been merged and the `maintenance/1.11.x` bra...

Co-authored-by: Kai <[email protected]>

Also move tabulations to from `__init__` to `__call__`, this
does not seem to matter much (can u...