Robust ratio-study breakdown binning + explicit bins option

[drussellmrichie]

Extracts breakdown bin-edge computation into a pure helper (_compute_breakdown_edges) and fixes two issues:

1. Robustness: the inline builder used np.quantile (returns NaN if the column has any NaN) and kept an edge whenever it was not in bins (no monotonicity guarantee). A heavily-skewed column (many rows sharing a value) collapsed the quantile edges to duplicates, so pd.cut raised "bins must increase monotonically" and crashed the entire ratio study. Now uses np.nanquantile, keeps only strictly-increasing edges, and returns no bins for a degenerate (all-NaN/constant) column so the breakdown is skipped instead of crashing.

2. New bins breakdown option: explicit value edges with optional bin_labels. quantiles give equal-count bins (degenerate on heavily-skewed columns) and slice_size formats labels as integers (unusable for 0–1 columns); explicit bins cover those cases.

Adds unit tests for the helper.

[github-actions]

Thank you for your contribution.
Please sign our CLA at the following link:
Click here to sign the CLA.

No action is required from you in this PR thread. Once you have signed the CLA externally, a maintainer will verify your signature and record it here on your behalf by commenting:

---

I affirm that this contributor has signed the CLA

---

Russell Richie seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You can retrigger this bot by commenting recheck in this Pull Request. Posted by the CLA Assistant Lite bot.}