Skip to content

Des/global refactoring#3

Merged
lilyreber merged 19 commits intomainfrom
des/global-refactoring
Apr 28, 2025
Merged

Des/global refactoring#3
lilyreber merged 19 commits intomainfrom
des/global-refactoring

Conversation

@Desiment
Copy link
Contributor

  • Fix mypy
  • Fix poetry
  • Clean-up
  • Renamed classes

Desiment added 19 commits April 27, 2025 00:23
project uses poetry, so this file actually not needed
ooetry needs only poetry.lock and pyproject.toml
we do not need pipenv, as we should use poetry
for packaging purposes (see poetry docs on this)
It seems to be a bug, that poetry does not allow for uppercase
and dash symbol in package name, (see related issue), so package
was renamed to `cfinversion`.

Related: python-poetry/poetry#8662
a few small changes:
* removed pytest from requirements (leaving it only in dev group)
* add jupyter-notebook dependency to dev group
Removed naive implementation of bohman methods. These implementations
do not satisfy interface (in type-hints), and do not use NumPy, so
they are quite slow, but have same precision.
It is not possible to implement Pareto distribution using only
scipy.special functions lib, because gammainc do not support
complex-valued argument. Removed incorrect implementation.
removed files for cauchy and lognormal distributions, as
they were empty
Renamed files related to inversion of charteristic functions to follow
PySATL-CPD style. Add imports to __init__.py files
* Renamed distribution related files to follow PySATL-CPD style.
* Rename Unif to Uniform
* Add squared uniform distribution
Value of S(t, x) = Imag[exp(itx) * phi(t) / (it)] is undefined at t = 0;
Given that phi'(0) = 0 (which is done via Standardizer), one can assume
that at t = 0 S(0, x) = -x, see https://arxiv.org/pdf/1701.08299 eq. 24
At poin t = 0 chr(t) should be 1, but previous implementation
encountered division by zero trying evaluate cf at this point.
Now, fit may take unstandardized cf, and cdf should work fine. Method
named ``cdf_`` reffers to implementation for standaridized variant
This was kind of useless, should add protocol instead
@Desiment Desiment requested a review from lilyreber April 28, 2025 15:25
@lilyreber lilyreber merged commit 2d3cc1e into main Apr 28, 2025
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants