Enable ruff B905 and require explicit zip strictness

.claude/CLAUDE.md

@@ -146,6 +146,19 @@ def estimate(self, y_true: NDArray, y_proxy: NDArray, power_tuning: bool = True)
 def estimate(self, y_true: NDArray, y_proxy: NDArray, *, power_tuning: bool = True) -> ...: ...
 ```
 
+### Zip Strictness
+
+Every `zip()` call must pass an explicit `strict=` argument (ruff `B905`). Use `strict=True` whenever the iterables are expected to have the same length, so a mismatch raises instead of silently truncating; use `strict=False` only when truncation is intended. This applies to library code, tests and notebook code cells alike.
+
+```python
+# BAD
+for estimate, label in zip(estimates, labels): ...
+
+
+# GOOD
+for estimate, label in zip(estimates, labels, strict=True): ...
+```
+
 ### Type Conversions
 
 Do not use needless type conversions like `float()` or `int()` unless required by the caller or for debugging purposes.

docs/deep_dive/case_studies/memerag.ipynb

@@ -338,7 +338,7 @@
     "fig, ax = plt.subplots(figsize=(11, 5.5))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates, strict=True):\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    for xc in [lo, hi]:\n",
     "        ax.plot([xc, xc], [y - 0.2, y + 0.2], color=color, linewidth=2.5, zorder=3)\n",
@@ -571,7 +571,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "glide-py (3.12.12.final.0)",
+   "display_name": "glide-py (3.12.12)",
    "language": "python",
    "name": "python3"
   },

docs/deep_dive/case_studies/spider.ipynb

@@ -231,7 +231,7 @@
     "fig, ax = plt.subplots(figsize=(11, 5.5))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates, strict=True):\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    for xc in [lo, hi]:\n",
     "        ax.plot([xc, xc], [y - 0.2, y + 0.2], color=color, linewidth=2.5, zorder=3)\n",

docs/deep_dive/scientific_validation/estimators/asi.ipynb

@@ -375,7 +375,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"ASI\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/clustered_ppi.ipynb

@@ -284,7 +284,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"Clustered PPI++\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/clustered_ptd.ipynb

@@ -284,7 +284,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"Clustered PTD\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/ipwptd.ipynb

@@ -375,7 +375,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"IPW PTD\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/multi_ppi.ipynb

@@ -304,7 +304,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"Multi-PPI++\": \"darkorange\", \"Proxy 1 only\": \"red\", \"Proxy 2 only\": \"firebrick\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/multi_ptd.ipynb

@@ -304,7 +304,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"Multi-PTD\": \"darkorange\", \"Proxy 1 only\": \"red\", \"Proxy 2 only\": \"firebrick\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/ppi.ipynb

@@ -278,7 +278,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"PPI++\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/ptd.ipynb

@@ -278,7 +278,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"PTD\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/stratified_ppi.ipynb

@@ -304,7 +304,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"Stratified PPI++\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/deep_dive/scientific_validation/estimators/stratified_ptd.ipynb

@@ -305,7 +305,7 @@
     "fig, axes = plt.subplots(1, 3, figsize=(15, 5), sharey=True)\n",
     "colors = {\"True only\": \"steelblue\", \"Stratified PTD\": \"darkorange\", \"Proxy only\": \"red\"}\n",
     "\n",
-    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels):\n",
+    "for ax, correlation, label in zip(axes, correlations_lmh, corr_labels, strict=True):\n",
     "    ax.plot(confidence_levels, confidence_levels, color=\"black\", lw=1.5, linestyle=\"--\", label=\"Ideal\")\n",
     "    for method in METHODS:\n",
     "        mean_ci = np.array([coverages_confidence_intervals[correlation][cl][method] for cl in confidence_levels])\n",

docs/tutorials/estimators/asi.ipynb

@@ -406,7 +406,7 @@
     "fig, ax = plt.subplots(figsize=(11, 5.5))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates, strict=True):\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    for xc in [lo, hi]:\n",
     "        ax.plot([xc, xc], [y - 0.2, y + 0.2], color=color, linewidth=2.5, zorder=3)\n",

docs/tutorials/estimators/clustered_ppi.ipynb

@@ -265,7 +265,7 @@
     "\n",
     "ax.bar(x_pos, bar_vals, color=bar_colors, width=0.45, edgecolor=\"white\", linewidth=2, zorder=3)\n",
     "\n",
-    "for xi, val, c in zip(x_pos, bar_vals, bar_colors):\n",
+    "for xi, val, c in zip(x_pos, bar_vals, bar_colors, strict=True):\n",
     "    ax.text(xi, val + 0.005, f\"{val:.1%}\", ha=\"center\", fontsize=14, fontweight=\"bold\", color=c)\n",
     "\n",
     "ax.annotate(\n",
@@ -390,7 +390,7 @@
     "fig, ax = plt.subplots(figsize=(11, 5.5))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates, strict=True):\n",
     "    # Confidence interval line\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    # Cap marks\n",

docs/tutorials/estimators/cost_optimal.ipynb

@@ -416,7 +416,7 @@
     "fig, ax = plt.subplots(figsize=(11, 5.5))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, baseline_estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, baseline_estimates, strict=True):\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    for xc in [lo, hi]:\n",
     "        ax.plot([xc, xc], [y - 0.2, y + 0.2], color=color, linewidth=2.5, zorder=3)\n",

docs/tutorials/estimators/cost_optimal_random.ipynb

@@ -409,7 +409,7 @@
     "fig, ax = plt.subplots(figsize=(11, 5.5))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, baseline_estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, baseline_estimates, strict=True):\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    for xc in [lo, hi]:\n",
     "        ax.plot([xc, xc], [y - 0.2, y + 0.2], color=color, linewidth=2.5, zorder=3)\n",

docs/tutorials/estimators/multi_ppi.ipynb

@@ -281,10 +281,10 @@
     "\n",
     "ax.bar(x_pos, bar_vals, color=bar_colors, width=0.45, edgecolor=\"white\", linewidth=2, zorder=3)\n",
     "\n",
-    "for xi_pos, val, c in zip(x_pos, bar_vals, bar_colors):\n",
+    "for xi_pos, val, c in zip(x_pos, bar_vals, bar_colors, strict=True):\n",
     "    ax.text(xi_pos, val + 0.004, f\"{val:.1%}\", ha=\"center\", fontsize=14, fontweight=\"bold\", color=c)\n",
     "\n",
-    "for xi_pos, bias in zip([0, 1], [gemini_bias, claude_bias]):\n",
+    "for xi_pos, bias in zip([0, 1], [gemini_bias, claude_bias], strict=True):\n",
     "    ax.text(\n",
     "        xi_pos,\n",
     "        t_mean + 0.055,\n",
@@ -410,7 +410,7 @@
     "fig, ax = plt.subplots(figsize=(11, 7))\n",
     "y_pos = [3, 2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates, strict=True):\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    for xc in [lo, hi]:\n",
     "        ax.plot([xc, xc], [y - 0.2, y + 0.2], color=color, linewidth=2.5, zorder=3)\n",

docs/tutorials/estimators/ppi.ipynb

@@ -258,7 +258,7 @@
     "\n",
     "ax.bar(x_pos, bar_vals, color=bar_colors, width=0.45, edgecolor=\"white\", linewidth=2, zorder=3)\n",
     "\n",
-    "for xi, val, c in zip(x_pos, bar_vals, bar_colors):\n",
+    "for xi, val, c in zip(x_pos, bar_vals, bar_colors, strict=True):\n",
     "    ax.text(xi, val + 0.005, f\"{val:.1%}\", ha=\"center\", fontsize=14, fontweight=\"bold\", color=c)\n",
     "\n",
     "ax.annotate(\n",
@@ -381,7 +381,7 @@
     "fig, ax = plt.subplots(figsize=(11, 5.5))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates, strict=True):\n",
     "    # Confidence interval line\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=4, solid_capstyle=\"round\", zorder=3)\n",
     "    # Cap marks\n",

docs/tutorials/estimators/stratified_ppi.ipynb

@@ -347,7 +347,7 @@
     ")\n",
     "\n",
     "# Annotate biases\n",
-    "for i, (j_mean, h_mean, bias) in enumerate(zip(judge_means, human_means, biases)):\n",
+    "for i, (j_mean, h_mean, bias) in enumerate(zip(judge_means, human_means, biases, strict=True)):\n",
     "    max_y = max(j_mean, h_mean)\n",
     "    ax.annotate(\n",
     "        \"\",\n",
@@ -473,7 +473,7 @@
     "fig, ax = plt.subplots(figsize=(12, 6))\n",
     "y_pos = [2, 1, 0]\n",
     "\n",
-    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates):\n",
+    "for y, (label, mean, lo, hi, color) in zip(y_pos, estimates, strict=True):\n",
     "    # Confidence interval line\n",
     "    ax.plot([lo, hi], [y, y], color=color, linewidth=5, solid_capstyle=\"round\", zorder=3)\n",
     "    # Cap marks\n",

glide/simulators/multi_binary.py

@@ -129,7 +129,8 @@ def generate_multi_binary_dataset(
     min_correlations = np.maximum(-p_t * p_p, p_p + p_t - 1 - p_t * p_p) / D
     max_correlations = np.minimum(p_t * (1 - p_p), p_p * (1 - p_t)) / D
 
-    for m, (rho, lo, hi, proxy_mean) in enumerate(zip(correlations_arr, min_correlations, max_correlations, p_p)):
+    bounds = zip(correlations_arr, min_correlations, max_correlations, p_p, strict=True)
+    for m, (rho, lo, hi, proxy_mean) in enumerate(bounds):
         if rho < lo or rho > hi:
             raise ValueError(
                 f"Proxy {m}: impossible combination of 'true_mean'={true_mean!r}, "

pyproject.toml

@@ -74,7 +74,7 @@ default-groups = ["dev"]
 [tool.ruff]
 line-length = 120
 fix = true
-lint.select = ["E", "F", "FBT001", "FBT002", "I", "RUF001", "RUF002", "RUF003", "D102", "D105", "B028"]
+lint.select = ["E", "F", "FBT001", "FBT002", "I", "RUF001", "RUF002", "RUF003", "D102", "D105", "B028", "B905"]
 
 [tool.ty.analysis]
 respect-type-ignore-comments = true

tests/unit/engines/test_clustered_ppi.py

@@ -65,7 +65,7 @@ def test_fit_tuning_parameter_delegates_to_ppi_engine(engine, dataset):
         tuning_parameter = engine._fit_tuning_parameter(dataset, power_tuning=False)
 
         mock_fit_tuning_parameter.assert_called_once()
-        for received, expected in zip(mock_fit_tuning_parameter.call_args[0][0], dataset):
+        for received, expected in zip(mock_fit_tuning_parameter.call_args[0][0], dataset, strict=True):
             np.testing.assert_array_equal(received, expected)
         assert mock_fit_tuning_parameter.call_args.kwargs == {"power_tuning": False}
         assert tuning_parameter == 0.5
@@ -80,7 +80,7 @@ def test_compute_mean_and_std_delegates_to_ppi_engine(engine, dataset):
         mean, std = engine._compute_mean_and_std(dataset, tuning_parameter=0.5)
 
         mock_compute_mean_and_std.assert_called_once()
-        for received, expected in zip(mock_compute_mean_and_std.call_args[0][0], dataset):
+        for received, expected in zip(mock_compute_mean_and_std.call_args[0][0], dataset, strict=True):
             np.testing.assert_array_equal(received, expected)
         assert mock_compute_mean_and_std.call_args[0][1] == 0.5
         assert (mean, std) == (1.5, 0.25)

tests/unit/simulators/test_batched_clustered_binary.py

@@ -56,7 +56,7 @@ def test_generate_batched_clustered_binary_dataset_delegates_to_validation(datas
 def test_generate_batched_clustered_binary_dataset_reproducibility(dataset_params):
     output1 = generate_batched_clustered_binary_dataset(**dataset_params, random_seed=42)
     output2 = generate_batched_clustered_binary_dataset(**dataset_params, random_seed=42)
-    for array1, array2 in zip(output1, output2):
+    for array1, array2 in zip(output1, output2, strict=True):
         np.testing.assert_array_equal(array1, array2)