{"spec_id":"gain-curve","library":"plotnine","language":"python","code":"\"\"\" anyplot.ai\ngain-curve: Cumulative Gains Chart\nLibrary: plotnine 0.15.4 | Python 3.13.13\nQuality: 92/100 | Updated: 2026-05-11\n\"\"\"\n\nimport os\n\nimport numpy as np\nimport pandas as pd\nfrom plotnine import (\n    aes,\n    element_line,\n    element_rect,\n    element_text,\n    geom_line,\n    ggplot,\n    labs,\n    scale_color_manual,\n    scale_x_continuous,\n    scale_y_continuous,\n    theme,\n    theme_minimal,\n)\n\n\n# Theme tokens\nTHEME = os.getenv(\"ANYPLOT_THEME\", \"light\")\nPAGE_BG = \"#FAF8F1\" if THEME == \"light\" else \"#1A1A17\"\nELEVATED_BG = \"#FFFDF6\" if THEME == \"light\" else \"#242420\"\nINK = \"#1A1A17\" if THEME == \"light\" else \"#F0EFE8\"\nINK_SOFT = \"#4A4A44\" if THEME == \"light\" else \"#B8B7B0\"\nINK_MUTED = \"#6B6A63\" if THEME == \"light\" else \"#A8A79F\"\n\n# Data - credit scoring: loan default classification model\nnp.random.seed(42)\nn_samples = 1000\n\n# Generate synthetic model scores and true labels\nbase_score = np.random.randn(n_samples)\n\n# Generate true labels (defaults) with ~15% default rate, correlated with score\nprob_default = 1 / (1 + np.exp(-(base_score + np.random.randn(n_samples) * 0.5)))\ny_true = (prob_default > 0.85).astype(int)\n\n# Model predictions - risk scores correlated with true labels but with noise\ny_score = base_score + np.where(y_true == 1, 1.2, -0.4) + np.random.randn(n_samples) * 0.7\ny_score = 1 / (1 + np.exp(-y_score))\n\n# Calculate cumulative gains curve\nsorted_indices = np.argsort(y_score)[::-1]\ny_true_sorted = y_true[sorted_indices]\n\ntotal_positives = np.sum(y_true)\ncumulative_positives = np.cumsum(y_true_sorted)\ngains = cumulative_positives / total_positives * 100\n\npercentages = np.arange(1, n_samples + 1) / n_samples * 100\n\n# Create DataFrame for plotting\ndf_model = pd.DataFrame({\"percent_population\": percentages, \"percent_positives\": gains, \"curve\": \"Model\"})\n\n# Random baseline (diagonal)\ndf_random = pd.DataFrame({\"percent_population\": [0, 100], \"percent_positives\": [0, 100], \"curve\": \"Random\"})\n\n# Perfect model: vertical rise to 100% at positive rate, then horizontal\npositive_rate = (total_positives / n_samples) * 100\ndf_perfect = pd.DataFrame(\n    {\"percent_population\": [0, positive_rate, 100], \"percent_positives\": [0, 100, 100], \"curve\": \"Perfect\"}\n)\n\n# Combine all curves\ndf = pd.concat([df_model, df_random, df_perfect], ignore_index=True)\n\n# Plot\nplot = (\n    ggplot(df, aes(x=\"percent_population\", y=\"percent_positives\", color=\"curve\"))\n    + geom_line(size=2.5)\n    + scale_color_manual(values={\"Model\": \"#009E73\", \"Random\": INK_MUTED, \"Perfect\": \"#AE3030\"})\n    + scale_x_continuous(breaks=range(0, 101, 20), limits=(0, 100))\n    + scale_y_continuous(breaks=range(0, 101, 20), limits=(0, 100))\n    + labs(\n        title=\"gain-curve · plotnine · anyplot.ai\",\n        x=\"Loan Portfolio Targeted (%)\",\n        y=\"Defaults Captured (%)\",\n        color=\"Curve\",\n    )\n    + theme_minimal()\n    + theme(\n        figure_size=(16, 9),\n        plot_background=element_rect(fill=PAGE_BG, color=PAGE_BG),\n        panel_background=element_rect(fill=PAGE_BG),\n        panel_grid_major=element_line(color=INK, size=0.3, alpha=0.10),\n        panel_grid_minor=element_line(color=INK, size=0.2, alpha=0.05),\n        panel_border=element_rect(color=INK_SOFT, fill=None, size=0.5),\n        axis_title=element_text(color=INK, size=20),\n        axis_text=element_text(color=INK_SOFT, size=16),\n        axis_line=element_line(color=INK_SOFT, size=0.5),\n        plot_title=element_text(color=INK, size=24),\n        legend_background=element_rect(fill=ELEVATED_BG, color=INK_SOFT),\n        legend_text=element_text(color=INK_SOFT, size=16),\n        legend_title=element_text(color=INK, size=16),\n    )\n)\n\n# Save\nplot.save(f\"plot-{THEME}.png\", dpi=300)\n"}