From b4d25c1370d319861f527e7929252ac3871c021c Mon Sep 17 00:00:00 2001
From: Andre Anjos <andre.dos.anjos@gmail.com>
Date: Wed, 3 Jul 2024 14:56:51 +0200
Subject: [PATCH] [libs.segmentation.engine.evaluator] Fix precision-recall
estimates when precision and recall == 0
---
src/mednet/libs/segmentation/engine/evaluator.py | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/src/mednet/libs/segmentation/engine/evaluator.py b/src/mednet/libs/segmentation/engine/evaluator.py
index 5e6ae0e5..027e876f 100644
--- a/src/mednet/libs/segmentation/engine/evaluator.py
+++ b/src/mednet/libs/segmentation/engine/evaluator.py
@@ -646,6 +646,11 @@ def run(
recall_curve = tpr_curve = numpy.array([recall(*k) for k in counts])
precision_curve = numpy.array([precision(*k) for k in counts])
+ # correction when precision is very small
+ precision_curve[
+ numpy.logical_and(precision_curve < 1e-8, recall_curve < 1e-8)
+ ] = 1.0
+
# populates data to be recorded in JSON format
eval_json_data.setdefault(split_name, {})["counts"] = {
k: v for k, v in zip(threshold_list, counts)
--
GitLab