diff --git a/README.rst b/README.rst index 73b9ac00d..4ad75c2b6 100644 --- a/README.rst +++ b/README.rst @@ -44,7 +44,7 @@ those documents into a versioned database that downstream users can query as a C workbook, or GeoPackage. .. image:: https://raw.githubusercontent.com/NatLabRockies/COMPASS/main/docs/source/_static/overview.png - :alt: High-level overview of the INFRA-COMPASS pipeline + :alt: High-level overview of the COMPASS pipeline :align: center :width: 100% @@ -64,17 +64,18 @@ architecture around the LLM call: jurisdictions. -Read more about the tool in the `documentation `_. +Read more about the tool in the `documentation `_, +or check out summaries of the validation results in the `validation report `_. Where is the extracted ordinance data? ====================================== -The National Laboratories of the Rockies (NLR) typically runs the INFRA-COMPASS pipeline +The National Laboratories of the Rockies (NLR) typically runs the COMPASS pipeline annually and publishes refreshed datasets to OpenEI. The latest published ordinance datasets are available here: -- Solar: https://data.openei.org/submissions/8519 -- Wind: https://data.openei.org/submissions/8602 +- Solar: https://data.openei.org/submissions/8602 +- Wind: https://data.openei.org/submissions/8519 Installing COMPASS diff --git a/docs/source/misc/about.rst b/docs/source/misc/about.rst index 3436acc80..1c3691f3e 100644 --- a/docs/source/misc/about.rst +++ b/docs/source/misc/about.rst @@ -162,3 +162,7 @@ See the development guides for full details: `plugin development `_ and `advanced plugin development `_. + +Validation +========== +Take a look at some high-level validation results in the `validation report `_. diff --git a/docs/source/val/geothermal_electricity_web_scraping_may_2026.png b/docs/source/val/geothermal_electricity_web_scraping_may_2026.png new file mode 100644 index 000000000..25c31d0c3 Binary files /dev/null and b/docs/source/val/geothermal_electricity_web_scraping_may_2026.png differ diff --git a/docs/source/val/solar_val_aug_2025.png b/docs/source/val/solar_val_aug_2025.png new file mode 100644 index 000000000..508942dd4 Binary files /dev/null and b/docs/source/val/solar_val_aug_2025.png differ diff --git a/docs/source/val/validation.rst b/docs/source/val/validation.rst index 357de92f1..3397bb4ec 100644 --- a/docs/source/val/validation.rst +++ b/docs/source/val/validation.rst @@ -5,6 +5,95 @@ Here we give a brief overview of the results of known COMPASS validation efforts COMPASS validation is an ongoing effort, and we will update this page as new results become available. +Geothermal Electricity Document Collection Validation (May 2026) +---------------------------------------------------------------- +This validation was for the document collection portion only. + +Info +^^^^ + +- **COMPASS Version**: `v0.15.2 `_ +- **Number of Documents**: 100 (Assuming 10,000 jurisdictions, there is a 95% chance that the metrics are within ±9.75% of the reported value) +- **Features**: None +- **Procedure Validated**: Document collection from web +- **LLM(s) used**: OpenAI GPT-4.1, OpenAI GPT-5 + + +Results +^^^^^^^ + +.. image:: geothermal_electricity_web_scraping_may_2026.png + + +Wind Document Collection Validation (September 2026) +---------------------------------------------------- +This validation was for the document collection portion only. + +Info +^^^^ + +- **COMPASS Version**: `v0.8.2 `_ +- **Number of Documents**: 100 (Assuming 10,000 jurisdictions, there is a 95% chance that the metrics are within ±9.75% of the reported value) +- **Features**: None +- **Procedure Validated**: Document collection from web +- **LLM(s) used**: OpenAI GPT-4o-mini, OpenAI GPT-4.1-nano, OpenAI GPT-4.1-mini, OpenAI GPT-4.1 + + +Results +^^^^^^^ + +.. image:: wind_web_scraping_september_2025.png + + +Solar Validation (August 2025) +------------------------------ + +This validation was for the ordinance value extraction portion only (assume documents are correct and belong +to the correct jurisdiction). + +This validation focused on the model ability to extract structured ordinance data from unstructured wind ordinance text documents. + + +Info +^^^^ + +- **COMPASS Version**: `v0.7.0 `_ +- **Number of Documents**: 78 (Assuming 10,000 jurisdictions, there is a 95% chance that the metrics are within ±11.05% of the reported value) +- **Features**: + + - structures (participating) + - property line (participating) + - structures (non-participating) + - property line (non-participating) + - roads + - railroads + - transmission + - water + - noise + - maximum height + - maximum project size + - minimum lot size + - maximum lot size + - density + - coverage + - prohibitions + - decommissioning + - glare + - visual impact + - primary use districts + - special use districts + - accessory use districts + - ordinance effective year + + +- **Procedure Validated**: Ordinance extraction from documents +- **LLM(s) used**: OpenAI GPT-4.1-nano, OpenAI GPT-4.1-mini, OpenAI GPT-4.1 + +Results +^^^^^^^ + +.. image:: solar_val_aug_2025.png + Mini Wind Validation (June 2025) -------------------------------- @@ -104,3 +193,5 @@ Results ^^^^^^^ .. image:: wind_jan_2024.png + +.. Margin of error calculator: https://www.calculator.net/sample-size-calculator.html diff --git a/docs/source/val/wind_web_scraping_september_2025.png b/docs/source/val/wind_web_scraping_september_2025.png new file mode 100644 index 000000000..c9d937b37 Binary files /dev/null and b/docs/source/val/wind_web_scraping_september_2025.png differ diff --git a/support/validation/May_2026_Geothermal_webscraping_validation.ipynb b/support/validation/May_2026_Geothermal_webscraping_validation.ipynb new file mode 100644 index 000000000..de105d925 --- /dev/null +++ b/support/validation/May_2026_Geothermal_webscraping_validation.ipynb @@ -0,0 +1,326 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "id": "edb8d2ec", + "metadata": {}, + "outputs": [], + "source": [ + "import pandas as pd\n", + "from val_utils import compute_stats\n", + "from val_utils import plot_compass_confusion_matrix_from_data\n", + "\n", + "import matplotlib\n", + "matplotlib.rcParams.update(matplotlib.rcParamsDefault)" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "c07828d1", + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "Who is correct? (Verified)\n", + "Both 62\n", + "Compass 28\n", + "Human 10\n", + "Name: count, dtype: int64" + ] + }, + "execution_count": 2, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "results = pd.read_csv(\"geothermal_webscraping_val_results.csv\")\n", + "results[\"Who is correct? (Verified)\"].value_counts()" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "28695d23", + "metadata": {}, + "outputs": [], + "source": [ + "def compute_scores(row, compass_scores, human_scores):\n", + " correctness = row[\"Who is correct? (Verified)\"]\n", + " ordinance_found = False\n", + " if correctness == \"Both\":\n", + " if row[\"Compass Link\"] == \"Not found\":\n", + " compass_scores[\"does not exist, not found\"] += 1\n", + " human_scores[\"does not exist, not found\"] += 1\n", + " else:\n", + " compass_scores[\"exists, found\"] += 1\n", + " human_scores[\"exists, found\"] += 1\n", + " ordinance_found = True\n", + " elif correctness == \"Compass\":\n", + " if row[\"Human Link\"] == \"Not found\": # COMPASS link must exist\n", + " compass_scores[\"exists, found\"] += 1\n", + " human_scores[\"exists, not found\"] += 1\n", + " ordinance_found = True\n", + " elif row[\"Compass Link\"] == \"Not found\": # human link exists\n", + " compass_scores[\"does not exist, not found\"] += 1\n", + " human_scores[\"does not exist, found\"] += 1\n", + " else: # both exist but COMPASS is correct, so human is incorrect\n", + " compass_scores[\"exists, found\"] += 1\n", + " human_scores[\"exists, incorrectly found\"] += 1\n", + " ordinance_found = True\n", + " elif correctness == \"Human\":\n", + " if row[\"Compass Link\"] == \"Not found\": # human link must exist\n", + " compass_scores[\"exists, not found\"] += 1\n", + " human_scores[\"exists, found\"] += 1\n", + " ordinance_found = True\n", + " elif row[\"Human Link\"] == \"Not found\": # COMPASS link exists\n", + " compass_scores[\"does not exist, found\"] += 1\n", + " human_scores[\"does not exist, not found\"] += 1\n", + " else: # both exist but human is correct, so COMPASS is incorrect\n", + " compass_scores[\"exists, incorrectly found\"] += 1\n", + " human_scores[\"exists, found\"] += 1\n", + " ordinance_found = True\n", + " elif correctness == \"Neither\":\n", + " if row[\"Compass Link\"] == \"Not found\":\n", + " if row[\"Human Link\"] == \"Not found\":\n", + " compass_scores[\"exists, not found\"] += 1\n", + " human_scores[\"exists, not found\"] += 1\n", + " ordinance_found = True\n", + " else: # guess here, assume human is more correct\n", + " compass_scores[\"exists, not found\"] += 1\n", + " human_scores[\"exists, incorrectly found\"] += 1\n", + " ordinance_found = True\n", + " elif row[\"Human Link\"] == \"Not found\":\n", + " compass_scores[\"exists, incorrectly found\"] += 1\n", + " human_scores[\"exists, not found\"] += 1\n", + " ordinance_found = True\n", + " else:\n", + " compass_scores[\"exists, incorrectly found\"] += 1\n", + " human_scores[\"exists, incorrectly found\"] += 1\n", + " ordinance_found = True\n", + " else:\n", + " msg = f\"Unexpected correctness value: {correctness}\"\n", + " raise ValueError(msg)\n", + "\n", + " return ordinance_found\n", + "\n", + "\n", + "def scores_dict_to_validation_df(scores, truth_labels_col=\"ordinance_exists\"):\n", + " label_to_truth = {\n", + " \"exists, found\": 1,\n", + " \"exists, not found\": 1,\n", + " \"exists, incorrectly found\": 0,\n", + " \"does not exist, found\": 0,\n", + " \"does not exist, not found\": 0,\n", + " }\n", + "\n", + " rows = []\n", + " for score_label, count in scores.items():\n", + " if score_label not in label_to_truth:\n", + " msg = f\"Unexpected score label: {score_label}\"\n", + " raise ValueError(msg)\n", + "\n", + " rows.extend(\n", + " {\n", + " \"Score\": score_label,\n", + " truth_labels_col: label_to_truth[score_label],\n", + " }\n", + " for _ in range(count)\n", + " )\n", + "\n", + " return pd.DataFrame(rows)" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "cec33a33", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Number of ordinances found: 45\n" + ] + } + ], + "source": [ + "compass_scores = {\n", + " \"exists, found\": 0,\n", + " \"exists, not found\": 0,\n", + " \"exists, incorrectly found\": 0,\n", + " \"does not exist, found\": 0,\n", + " \"does not exist, not found\": 0\n", + "}\n", + "human_scores = {\n", + " \"exists, found\": 0,\n", + " \"exists, not found\": 0,\n", + " \"exists, incorrectly found\": 0,\n", + " \"does not exist, found\": 0,\n", + " \"does not exist, not found\": 0\n", + "}\n", + "ordinance_found = 0\n", + "\n", + "for __, row in results.iterrows():\n", + " ordinance_found += compute_scores(row, compass_scores, human_scores)\n", + "\n", + "assert sum(compass_scores.values()) == len(results)\n", + "assert sum(human_scores.values()) == len(results)\n", + "print(\"Number of ordinances found:\", ordinance_found)" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "1d33dd99", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "COMPASS\n", + "confusion matrix:\n", + " [[36 4]\n", + " [ 1 59]]\n", + "accuracy=90.00%, precision=85.71%, recall=90.00%\n" + ] + } + ], + "source": [ + "score_cats = {\n", + " \"exists_report\": {\"exists, found\"},\n", + " \"exists_no_report\": {\"exists, not found\"},\n", + " \"exists_bad_report\": {\"exists, incorrectly found\"},\n", + " \"dne_report\": {\"does not exist, found\"},\n", + " \"dne_no_report\": {\"does not exist, not found\"},\n", + "}\n", + "\n", + "truth_labels_col = \"ordinance_exists\"\n", + "\n", + "compass_df = scores_dict_to_validation_df(\n", + " compass_scores,\n", + " truth_labels_col=truth_labels_col,\n", + ")\n", + "compass_cm, compass_acc, compass_precision, compass_recall = compute_stats(\n", + " compass_df,\n", + " score_cats,\n", + " truth_labels_col=truth_labels_col,\n", + ")\n", + "\n", + "\n", + "print(\"COMPASS\")\n", + "print(\"confusion matrix:\\n\", compass_cm)\n", + "print(\n", + " f\"accuracy={compass_acc:.2%}, \"\n", + " f\"precision={compass_precision:.2%}, \"\n", + " f\"recall={compass_recall:.2%}\"\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "id": "05d4887c", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plot_compass_confusion_matrix_from_data(\n", + " compass_df,\n", + " score_cats,\n", + " title=\"COMPASS Geothermal Webscraping Validation\",\n", + " truth_labels_col=truth_labels_col,\n", + " x_label=\"COMPASS Retrieved Ordinance\",\n", + " y_label=\"Ordinance Exists\",\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "be6ff130", + "metadata": {}, + "outputs": [], + "source": [ + "human_df = scores_dict_to_validation_df(\n", + " human_scores,\n", + " truth_labels_col=truth_labels_col,\n", + ")\n", + "human_cm, human_acc, human_precision, human_recall = compute_stats(\n", + " human_df,\n", + " score_cats,\n", + " truth_labels_col=truth_labels_col,\n", + ")\n", + "\n", + "print(\"\\nHUMAN\")\n", + "print(\"confusion matrix:\\n\", human_cm)\n", + "print(\n", + " f\"accuracy={human_acc:.2%}, \"\n", + " f\"precision={human_precision:.2%}, \"\n", + " f\"recall={human_recall:.2%}\"\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "c362b67c", + "metadata": {}, + "outputs": [], + "source": [ + "plot_compass_confusion_matrix_from_data(\n", + " human_df,\n", + " score_cats,\n", + " title=\"Human Geothermal Webscraping Validation\",\n", + " truth_labels_col=truth_labels_col,\n", + " x_label=\"Human Retrieved Ordinance\",\n", + " y_label=\"Ordinance Exists\",\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "dfa1877a", + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.12.13" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/support/compute_validation_metrics.ipynb b/support/validation/compute_validation_metrics-general.ipynb similarity index 100% rename from support/compute_validation_metrics.ipynb rename to support/validation/compute_validation_metrics-general.ipynb diff --git a/support/val_utils.py b/support/validation/val_utils.py similarity index 100% rename from support/val_utils.py rename to support/validation/val_utils.py