Hi Kai, I am not using numpy financial, but from what I know: Given numpy 2.0 has a support for MPFD types I think it is very reasonable to drop Decimal support and remove unnecessary complexity. Regards, dgpb

On 30 Mar 2024, at 02:20, kaistriega@gmail.com wrote:

Hi all,

I'm currently working NumPy-Financial (npf) and would like to ask for comments on dropping decimal.Decimal support in npf. I'm proposing to drop Decimal support for npf. There are a few reasons for this:

1. No one appears to use decimal types. Since it's release in 2019, not a single issue has been raised that uses decimal support. Furthermore, I have taken a quick survey of dependent projects, none of which seems to use decimals.

2. It's not actually supported that thoroughly. Npf claims that: “Functions support the :class:`decimal.Decimal` type unless otherwise stated.”. This isn't technically true. Many functions that claim to support decimal classes don't. I think it would be better to have a consistent interface that would either entierly support decimal or not support it at all.

3. I'm currently working on mimicking broadcasting behaviour for functions. This involves lots of for loops. To speed these up I will use either numba or Cython. Neither of these support decimal classes natively.

If you have any comments on whether this seems to be the right option, I would be very interested in hearing it.

Kind regards

Kai _______________________________________________ NumPy-Discussion mailing list -- numpy-discussion@python.org To unsubscribe send an email to numpy-discussion-leave@python.org https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: dom.grigonis@gmail.com