update recipe to use numpy-base

[vtavana]

This will help to have the following dependency mkl_* -> numpy-base and then numpy -> mkl_*, numpy-base.

[Copilot]

Pull Request Overview

Updates the conda recipe to use numpy-base instead of numpy for dependency management. This change establishes a cleaner dependency hierarchy where MKL packages depend on numpy-base and numpy depends on both MKL packages and numpy-base.

• Replace numpy with numpy-base in the recipe's runtime dependencies
• Modify CI workflows to handle scipy installation conditionally for Python 3.9 due to dependency conflicts
• Remove Python 3.9-specific MKL-devel version constraint

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
conda-recipe/meta.yaml	Updates runtime dependency from numpy to numpy-base and removes Python 3.9 build constraint
.github/workflows/conda-package.yml	Modifies test workflows to install scipy conditionally to avoid dependency conflicts with Python 3.9