[codex] Respect zero probability mutation

[alerodriargui]

Description

Fixes the probability mutation defaulting logic so mutation_prob=0 is respected instead of being treated as omitted.

Previously, Chromosome.mutate() used a truthiness check, so passing 0 fell back to the default 1 / len(self.data) probability and could still mutate genes. The new check defaults only when mutation_prob is None.

Adds a regression test covering the zero-probability case.

Related issue

Closes #14

Type of change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that changes existing behavior)
• Documentation only

Checklist

• I have signed the CLA (the CLA Assistant bot will prompt external contributors)
• My commit messages follow Conventional Commits
• ruff check . and black --check . pass
• mypy genetic_algorithm passes
• pytest passes
• The engine remains dependency-free (standard library only)
• I have added/updated tests where relevant
• I have updated documentation where relevant
• No secrets, API keys, internal URLs, or proprietary content are included

[github-actions]

All contributors have signed the CLA ✍️ ✅
_{Posted by the CLA Assistant Lite bot.}

[opensource-SantanderAI]

Hi @alerodriargui — thank you for this contribution!

We reviewed PR #15 ("Respect zero probability mutation"): the is not None fix and the added regression test are spot-on, and all CI checks are green except the CLA.

Signing the Contributor License Agreement is a mandatory step before we can merge any external contribution, and it takes about 30 seconds — no forms. Please post a new comment on this PR with exactly this text:

I have read the CLA Document and I hereby sign the CLA

• CLA document: https://github.com/SantanderAI/cla/blob/main/CLA.md

As soon as the CLA Assistant re-runs and turns green, we'll merge right away. Thanks again!

[alerodriargui]

I have read the CLA Document and I hereby sign the CLA

[opensource-SantanderAI]

Approved. Correct upstream fix: mutation now uses an explicit is not None check so a configured mutation probability of 0.0 is respected instead of falling back to the default (the previous falsy check treated 0.0 as unset). The added regression test covers the zero-probability case. CLA is now signed and CI is green. Thanks @alerodriargui!