BUG FIX: scalar Pow.impl returning complex for fractional powers of negative floats

[vybhav72954]

Description

Pow.impl in pytensor/scalar/basic.py uses Python's native ** operator, which returns complex numbers for fractional powers of negative floats (e.g., (-1.0) ** 0.01 returns (0.999+0.031j)). This is inconsistent with the tensor path, which already uses np.power via nfunc_spec and correctly returns nan.

The complex result then crashes _cast_to_promised_scalar_dtype when it tries to cast to float64.

Fix: Replace x**y with np.power(x, y) in Pow.impl, making the scalar path consistent with the tensor path.

Related Issue

• Closes BUG: Scalar Pow.impl returns complex instead of nan for fractional powers of negative floats #1944
• Related to #

Checklist

• Checked that the pre-commit linting/style checks pass
• Included tests that prove the fix is effective or that the new feature works
• Added necessary documentation (docstrings and/or example notebooks)
• If you are a pro: each commit corresponds to a relevant logical change

Type of change

• New feature / enhancement
• Bug fix
• Documentation
• Maintenance
• Other (please specify):