feat(highcharts): implement lift-curve (#2414)

github-actions[bot] · web-flow · commit a22071527f18 · 2025-12-27T19:36:39.000Z
## Implementation: `lift-curve` - highcharts Implements the **highcharts** version of `lift-curve`. **File:** `plots/lift-curve/implementations/highcharts.py` --- :robot: *[impl-generate workflow](https://github.com/MarkusNeusinger/pyplots/actions/runs/20543283913)* --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
diff --git a/plots/lift-curve/implementations/highcharts.py b/plots/lift-curve/implementations/highcharts.py
@@ -0,0 +1,243 @@
+""" pyplots.ai
+lift-curve: Model Lift Chart
+Library: highcharts unknown | Python 3.13.11
+Quality: 91/100 | Created: 2025-12-27
+"""
+
+import tempfile
+import time
+import urllib.request
+from pathlib import Path
+
+import numpy as np
+from highcharts_core.chart import Chart
+from highcharts_core.options import HighchartsOptions
+from highcharts_core.options.series.area import LineSeries
+from highcharts_core.options.series.scatter import ScatterSeries
+from selenium import webdriver
+from selenium.webdriver.chrome.options import Options
+
+
+# Data: Simulated customer response prediction
+np.random.seed(42)
+n_samples = 1000
+
+# Generate realistic model scores and outcomes
+# Good model: higher scores correlate with positive outcomes
+scores = np.random.beta(2, 5, n_samples)  # Model probability scores
+noise = np.random.random(n_samples)
+# True positives more likely for higher scores
+y_true = (scores + 0.3 * noise > 0.35).astype(int)
+
+# Calculate lift curve data
+sorted_indices = np.argsort(scores)[::-1]  # Sort by score descending
+y_true_sorted = y_true[sorted_indices]
+
+# Calculate cumulative lift at each percentile
+n_positive = y_true.sum()
+baseline_rate = n_positive / n_samples
+cumulative_positives = np.cumsum(y_true_sorted)
+population_pct = np.arange(1, n_samples + 1) / n_samples * 100
+
+# Lift = (cumulative response rate) / (baseline response rate)
+cumulative_response_rate = cumulative_positives / np.arange(1, n_samples + 1)
+lift = cumulative_response_rate / baseline_rate
+
+# Sample at regular intervals for smooth curve (every 1%)
+sample_points = list(range(0, n_samples, max(1, n_samples // 100)))
+if sample_points[-1] != n_samples - 1:
+    sample_points.append(n_samples - 1)
+
+pct_sampled = [population_pct[i] for i in sample_points]
+lift_sampled = [float(lift[i]) for i in sample_points]
+
+# Create chart
+chart = Chart(container="container")
+chart.options = HighchartsOptions()
+
+# Chart configuration with proper margins
+chart.options.chart = {
+    "type": "line",
+    "width": 4800,
+    "height": 2700,
+    "backgroundColor": "#ffffff",
+    "marginBottom": 200,
+    "marginLeft": 220,
+    "marginRight": 150,
+    "marginTop": 180,
+    "spacingBottom": 50,
+}
+
+# Title
+chart.options.title = {
+    "text": "lift-curve \u00b7 highcharts \u00b7 pyplots.ai",
+    "style": {"fontSize": "56px", "fontWeight": "bold"},
+    "y": 60,
+}
+
+# Subtitle with model info
+chart.options.subtitle = {
+    "text": f"Customer Response Model | Baseline Rate: {baseline_rate:.1%}",
+    "style": {"fontSize": "36px", "color": "#666666"},
+    "y": 110,
+}
+
+# X-axis
+chart.options.x_axis = {
+    "title": {"text": "Population Targeted (%)", "style": {"fontSize": "40px", "fontWeight": "bold"}, "margin": 30},
+    "labels": {"style": {"fontSize": "32px"}},
+    "min": 0,
+    "max": 100,
+    "tickInterval": 10,
+    "gridLineWidth": 1,
+    "gridLineColor": "#e0e0e0",
+    "lineWidth": 2,
+}
+
+# Y-axis - start from 0.9 to better show the lift curve variation
+chart.options.y_axis = {
+    "title": {"text": "Cumulative Lift", "style": {"fontSize": "40px", "fontWeight": "bold"}, "margin": 30},
+    "labels": {"style": {"fontSize": "32px"}},
+    "min": 0.9,
+    "max": 1.6,
+    "tickInterval": 0.1,
+    "gridLineWidth": 1,
+    "gridLineColor": "#e0e0e0",
+    "plotLines": [
+        {
+            "value": 1,
+            "color": "#888888",
+            "width": 4,
+            "dashStyle": "Dash",
+            "zIndex": 5,
+            "label": {
+                "text": "Random Selection (Lift = 1)",
+                "style": {"fontSize": "28px", "color": "#666666", "fontWeight": "bold"},
+                "align": "right",
+                "x": -20,
+                "y": 20,
+            },
+        }
+    ],
+}
+
+# Legend
+chart.options.legend = {
+    "enabled": True,
+    "itemStyle": {"fontSize": "32px"},
+    "align": "right",
+    "verticalAlign": "top",
+    "layout": "vertical",
+    "x": -30,
+    "y": 100,
+    "symbolRadius": 6,
+    "symbolWidth": 30,
+    "symbolHeight": 16,
+}
+
+# Plot options
+chart.options.plot_options = {"line": {"lineWidth": 6, "marker": {"enabled": False}}, "series": {"animation": False}}
+
+# Create lift curve series
+lift_series = LineSeries()
+lift_series.name = "Model Lift"
+lift_series.data = [[pct_sampled[i], lift_sampled[i]] for i in range(len(pct_sampled))]
+lift_series.color = "#306998"  # Python Blue
+lift_series.marker = {"enabled": True, "radius": 8, "symbol": "circle"}
+lift_series.lineWidth = 6
+
+chart.add_series(lift_series)
+
+# Add key decile annotations as a single series with data labels
+decile_points = [10, 20, 30, 50]
+annotation_data = []
+for pct in decile_points:
+    idx = min(int(pct * n_samples / 100) - 1, n_samples - 1)
+    lift_val = lift[idx]
+    annotation_data.append(
+        {
+            "x": pct,
+            "y": float(lift_val),
+            "dataLabels": {
+                "enabled": True,
+                "format": f"{lift_val:.2f}x",
+                "style": {"fontSize": "28px", "fontWeight": "bold", "color": "#306998"},
+                "y": -30,
+                "backgroundColor": "rgba(255, 255, 255, 0.8)",
+                "borderRadius": 5,
+                "padding": 8,
+            },
+        }
+    )
+
+annotation_series = ScatterSeries()
+annotation_series.name = "Key Percentiles"
+annotation_series.data = annotation_data
+annotation_series.color = "#FFD43B"  # Python Yellow
+annotation_series.marker = {
+    "enabled": True,
+    "radius": 14,
+    "symbol": "diamond",
+    "fillColor": "#FFD43B",
+    "lineColor": "#306998",
+    "lineWidth": 3,
+}
+
+chart.add_series(annotation_series)
+
+# Download Highcharts JS for inline embedding
+highcharts_url = "https://code.highcharts.com/highcharts.js"
+with urllib.request.urlopen(highcharts_url, timeout=30) as response:
+    highcharts_js = response.read().decode("utf-8")
+
+# Get the chart options as JS - avoid DOMContentLoaded wrapper issue
+js_literal = chart.to_js_literal()
+# Remove the DOMContentLoaded wrapper if present
+if "DOMContentLoaded" in js_literal:
+    # Extract just the Highcharts.chart(...) call
+    import re
+
+    match = re.search(r"(Highcharts\.chart\([^;]+\));", js_literal, re.DOTALL)
+    if match:
+        js_literal = match.group(1) + ";"
+
+# Generate HTML with inline scripts - use window.onload for reliable rendering
+html_content = f"""<!DOCTYPE html>
+<html>
+<head>
+    <meta charset="utf-8">
+    <script>{highcharts_js}</script>
+</head>
+<body style="margin:0;">
+    <div id="container" style="width: 4800px; height: 2700px;"></div>
+    <script>
+    window.onload = function() {{
+        {js_literal}
+    }};
+    </script>
+</body>
+</html>"""
+
+# Save HTML version
+with open("plot.html", "w", encoding="utf-8") as f:
+    f.write(html_content)
+
+# Export to PNG via Selenium
+with tempfile.NamedTemporaryFile(mode="w", suffix=".html", delete=False, encoding="utf-8") as f:
+    f.write(html_content)
+    temp_path = f.name
+
+chrome_options = Options()
+chrome_options.add_argument("--headless")
+chrome_options.add_argument("--no-sandbox")
+chrome_options.add_argument("--disable-dev-shm-usage")
+chrome_options.add_argument("--disable-gpu")
+chrome_options.add_argument("--window-size=4800,2700")
+
+driver = webdriver.Chrome(options=chrome_options)
+driver.get(f"file://{temp_path}")
+time.sleep(5)  # Wait for chart to render
+driver.save_screenshot("plot.png")
+driver.quit()
+
+Path(temp_path).unlink()
diff --git a/plots/lift-curve/metadata/highcharts.yaml b/plots/lift-curve/metadata/highcharts.yaml
@@ -0,0 +1,27 @@
+library: highcharts
+specification_id: lift-curve
+created: '2025-12-27T19:24:09Z'
+updated: '2025-12-27T19:33:31Z'
+generated_by: claude-opus-4-5-20251101
+workflow_run: 20543283913
+issue: 0
+python_version: 3.13.11
+library_version: unknown
+preview_url: https://storage.googleapis.com/pyplots-images/plots/lift-curve/highcharts/plot.png
+preview_thumb: https://storage.googleapis.com/pyplots-images/plots/lift-curve/highcharts/plot_thumb.png
+preview_html: https://storage.googleapis.com/pyplots-images/plots/lift-curve/highcharts/plot.html
+quality_score: 91
+review:
+  strengths:
+  - Excellent use of Highcharts plotLines feature for the reference line with label
+  - Effective combination of LineSeries for the curve and ScatterSeries for key percentile
+    annotations
+  - Clean colorblind-safe palette (Python blue/yellow)
+  - Proper title format and informative subtitle with baseline rate
+  - Good data labels on key percentiles showing lift multipliers
+  weaknesses:
+  - The Random Selection (Lift = 1) label appears to be cut off at the right edge
+    of the plot
+  - Lift values at 10%, 20%, and 30% are identical (1.52x), which looks artificial;
+    the data generation could show more variation
+  - Code complexity with regex extraction of JS literal could be simplified
