Make StudentT/AsymmetricLaplace .dist() signatures consistent

[cluhmann]

This PR closes #5602. Breaking changes:

• StudentT dist signature changed to match Normal, HalfStudentT, and others
• StudentT now requires either sigma or lam as kwarg to alert uses to change
• StudentT's nu argument now defaults to 1.0 (now matches HalfStudentT)
• HalfStudentT's nu argument now has no default (now matches StudentT)
• AsymmetricLaplace dist signature changed to match Laplace, SkewNormal, and StudentT
• AsymmetricLaplace's mu argument no longer has a default value (now matches Laplace)

[codecov]

Codecov Report

Merging #5628 (f33fc00) into main (fa015e3) will increase coverage by 1.74%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #5628      +/-   ##
==========================================
+ Coverage   86.86%   88.61%   +1.74%     
==========================================
  Files          75       75              
  Lines       13715    13715              
==========================================
+ Hits        11914    12153     +239     
+ Misses       1801     1562     -239     

Impacted Files	Coverage Δ
pymc/distributions/continuous.py	98.18% <100.00%> (ø)
pymc/sampling.py	88.15% <0.00%> (+0.11%)	⬆️
pymc/distributions/multivariate.py	92.15% <0.00%> (+0.12%)	⬆️
pymc/model.py	86.02% <0.00%> (+0.41%)	⬆️
pymc/distributions/logprob.py	96.33% <0.00%> (+0.91%)	⬆️
pymc/util.py	76.25% <0.00%> (+1.87%)	⬆️
pymc/distributions/distribution.py	90.80% <0.00%> (+2.29%)	⬆️
pymc/aesaraf.py	91.20% <0.00%> (+2.76%)	⬆️
pymc/backends/ndarray.py	79.64% <0.00%> (+7.07%)	⬆️
pymc/tuning/starting.py	92.68% <0.00%> (+12.19%)	⬆️
... and 6 more

[ricardoV94]

Will need release-notes as well

[cluhmann]

Should the signatures of get_moment() and logp() match the signature of dist()? I had assumed that the former were used mostly internally, but looking again they seem to be reasonably public. Should have asked earlier.

[ricardoV94]

Should the signatures of get_moment() and logp() match the signature of dist()? I had assumed that the former were used mostly internally, but looking again they seem to be reasonably public. Should have asked earlier.

No, those should be considered internal, and the order is fixed by the RandomVariable

[ricardoV94]

One HalfStudentT test seems to be failing. It probably relied on old parameter order or default, and has to be updated: https://github.com/pymc-devs/pymc/runs/5632254063?check_suite_focus=true

[OriolAbril]

No, those should be considered internal, and the order is fixed by the RandomVariable

And these methods should be removed from docs. The only reason they are not is the method listing in classes is automatic. With @ricardoV94 we tried a template to list only (but always) the dist method, but some distributions are actually functions/wrappers that don't have a dist and we weren't able to figure out conditionals in autodoc templates, not sure if we have an issue for that yet though

[ricardoV94]

Suggested change

	- `pm.StudentT` now now requires either `sigma` or `lam` as kwarg [#5628](https://github.com/pymc-devs/pymc/pull/5628)
	- `pm.StudentT` now requires `nu` to be specified (no longer defaults to 1)
	- `pm.AsymmetricLaplace` positional arguments re-ordered
	- `pm.AsymmetricLaplace` now requires `mu` to be specified (no longer defaults to 0)
	Signature and default parameters changed for several distributions (see [#5628](https://github.com/pymc-devs/pymc/pull/5628)):
	- `pm.StudentT` now requires either `sigma` or `lam` as kwarg
	- `pm.StudentT` now requires `nu` to be specified (no longer defaults to 1)
	- `pm.AsymmetricLaplace` positional arguments re-ordered
	- `pm.AsymmetricLaplace` now requires `mu` to be specified (no longer defaults to 0)

Also removes a double now

[ricardoV94]

This change shouldn't be in this PR. You might need to rebase from main