Jens/geointel
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

audit: seal accuracy phase 1 evidence

Browse Source
This commit is contained in:
Jens
2026-08-01 19:02:24 +02:00
parent 3d442ef43f
commit cebeb5f3b4
29 changed files with 11269 additions and 10990 deletions
Download Patch File Download Diff File Expand all files Collapse all files
artifacts/evidence/accuracy/P1/golden-qa-run-1.json
+177 -177
View File
@@ -1,177 +1,177 @@
{
"benchmark_id": "golden-buildings-partial-match-v1",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings.geojson",
"reference": "fixtures/golden/reference_buildings.geojson"
},
"metrics": {
"f1": 0.5,
"false_negative_count": 1,
"false_positive_count": 1,
"mean_iou": 0.8339768339652592,
"precision": 0.5,
"recall": 0.5
},
"persistence": {
"commit_count": 4,
"metric_count": 24,
"metric_keys": [
"f1",
"false_negative_count",
"false_positive_count",
"mean_iou",
"precision",
"recall"
],
"quality_check_count": 4
},
"scenario_count": 4,
"scenarios": [
{
"benchmark_id": "golden-buildings-partial-match-v1",
"description": "Two reference building polygons and two candidate building polygons: one candidate matches ref-1, one candidate is a false positive, and ref-2 is a false negative.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings.geojson",
"reference": "fixtures/golden/reference_buildings.geojson"
},
"metrics": {
"f1": 0.5,
"false_negative_count": 1,
"false_positive_count": 1,
"mean_iou": 0.8339768339652592,
"precision": 0.5,
"recall": 0.5
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "5fe92da9-2d8d-46a4-9992-2de89e3cf511",
"result_counts": {
"candidate_feature_count": 2,
"matches": 1,
"reference_feature_count": 2
},
"status": "passed"
},
{
"benchmark_id": "golden-buildings-perfect-match-v1",
"description": "One candidate building polygon exactly matches one reference polygon.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings_perfect.geojson",
"reference": "fixtures/golden/reference_buildings_perfect.geojson"
},
"metrics": {
"f1": 1.0,
"false_negative_count": 0,
"false_positive_count": 0,
"mean_iou": 1.0,
"precision": 1.0,
"recall": 1.0
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "7b8c1444-31e6-4780-92cb-f2060f4a4048",
"result_counts": {
"candidate_feature_count": 1,
"matches": 1,
"reference_feature_count": 1
},
"status": "passed"
},
{
"benchmark_id": "golden-buildings-no-overlap-v1",
"description": "One candidate polygon and one reference polygon do not overlap.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings_no_overlap.geojson",
"reference": "fixtures/golden/reference_buildings_no_overlap.geojson"
},
"metrics": {
"f1": null,
"false_negative_count": 1,
"false_positive_count": 1,
"mean_iou": null,
"precision": 0.0,
"recall": 0.0
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "1ffb1817-b237-4ca8-aad7-ff3312ac7e03",
"result_counts": {
"candidate_feature_count": 1,
"matches": 0,
"reference_feature_count": 1
},
"status": "passed"
},
{
"benchmark_id": "golden-buildings-multipolygon-match-v1",
"description": "One candidate MultiPolygon exactly matches one reference MultiPolygon.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings_multipolygon.geojson",
"reference": "fixtures/golden/reference_buildings_multipolygon.geojson"
},
"metrics": {
"f1": 1.0,
"false_negative_count": 0,
"false_positive_count": 0,
"mean_iou": 1.0,
"precision": 1.0,
"recall": 1.0
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "60e6a321-2aaf-467b-9281-f30dd4179333",
"result_counts": {
"candidate_feature_count": 1,
"matches": 1,
"reference_feature_count": 1
},
"status": "passed"
}
],
"status": "passed",
"version": 1
}
{
"benchmark_id": "golden-buildings-partial-match-v1",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings.geojson",
"reference": "fixtures/golden/reference_buildings.geojson"
},
"metrics": {
"f1": 0.5,
"false_negative_count": 1,
"false_positive_count": 1,
"mean_iou": 0.8339768339652592,
"precision": 0.5,
"recall": 0.5
},
"persistence": {
"commit_count": 4,
"metric_count": 24,
"metric_keys": [
"f1",
"false_negative_count",
"false_positive_count",
"mean_iou",
"precision",
"recall"
],
"quality_check_count": 4
},
"scenario_count": 4,
"scenarios": [
{
"benchmark_id": "golden-buildings-partial-match-v1",
"description": "Two reference building polygons and two candidate building polygons: one candidate matches ref-1, one candidate is a false positive, and ref-2 is a false negative.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings.geojson",
"reference": "fixtures/golden/reference_buildings.geojson"
},
"metrics": {
"f1": 0.5,
"false_negative_count": 1,
"false_positive_count": 1,
"mean_iou": 0.8339768339652592,
"precision": 0.5,
"recall": 0.5
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "5fe92da9-2d8d-46a4-9992-2de89e3cf511",
"result_counts": {
"candidate_feature_count": 2,
"matches": 1,
"reference_feature_count": 2
},
"status": "passed"
},
{
"benchmark_id": "golden-buildings-perfect-match-v1",
"description": "One candidate building polygon exactly matches one reference polygon.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings_perfect.geojson",
"reference": "fixtures/golden/reference_buildings_perfect.geojson"
},
"metrics": {
"f1": 1.0,
"false_negative_count": 0,
"false_positive_count": 0,
"mean_iou": 1.0,
"precision": 1.0,
"recall": 1.0
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "7b8c1444-31e6-4780-92cb-f2060f4a4048",
"result_counts": {
"candidate_feature_count": 1,
"matches": 1,
"reference_feature_count": 1
},
"status": "passed"
},
{
"benchmark_id": "golden-buildings-no-overlap-v1",
"description": "One candidate polygon and one reference polygon do not overlap.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings_no_overlap.geojson",
"reference": "fixtures/golden/reference_buildings_no_overlap.geojson"
},
"metrics": {
"f1": null,
"false_negative_count": 1,
"false_positive_count": 1,
"mean_iou": null,
"precision": 0.0,
"recall": 0.0
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "1ffb1817-b237-4ca8-aad7-ff3312ac7e03",
"result_counts": {
"candidate_feature_count": 1,
"matches": 0,
"reference_feature_count": 1
},
"status": "passed"
},
{
"benchmark_id": "golden-buildings-multipolygon-match-v1",
"description": "One candidate MultiPolygon exactly matches one reference MultiPolygon.",
"fixtures": {
"candidate": "fixtures/golden/predicted_buildings_multipolygon.geojson",
"reference": "fixtures/golden/reference_buildings_multipolygon.geojson"
},
"metrics": {
"f1": 1.0,
"false_negative_count": 0,
"false_positive_count": 0,
"mean_iou": 1.0,
"precision": 1.0,
"recall": 1.0
},
"persistence": {
"commit_count": 1,
"metric_count": 6,
"metric_keys": [
"precision",
"recall",
"f1",
"mean_iou",
"false_positive_count",
"false_negative_count"
],
"quality_check_count": 1
},
"quality_check_id": "60e6a321-2aaf-467b-9281-f30dd4179333",
"result_counts": {
"candidate_feature_count": 1,
"matches": 1,
"reference_feature_count": 1
},
"status": "passed"
}
],
"status": "passed",
"version": 1
}