update

pipline_test/display_result.py

@@ -0,0 +1,144 @@
+from sklearn.preprocessing import label_binarize
+from sklearn.metrics import precision_recall_curve, average_precision_score
+import matplotlib.pyplot as plt
+import numpy as np
+# from utils.config import conf
+import os
+
+
+# def show_multiclass_pr(y_true, y_scores):  # 多分类
+#     # 将真实标签二值化
+#     y_true_binarized = label_binarize(y_true, classes=np.arange(conf.n_classes))
+#
+#     # 计算每个类别的PR曲线和平均精度
+#     precision = dict()
+#     recall = dict()
+#     average_precision = dict()
+#
+#     for i in range(conf.n_classes):
+#         precision[i], recall[i], _ = precision_recall_curve(y_true_binarized[:, i], y_scores[:, i])
+#         average_precision[i] = average_precision_score(y_true_binarized[:, i], y_scores[:, i])
+#
+#     # 计算微平均PR曲线和平均精度
+#     precision["micro"], recall["micro"], _ = precision_recall_curve(y_true_binarized.ravel(), y_scores.ravel())
+#     average_precision["micro"] = average_precision_score(y_true_binarized, y_scores, average="micro")
+#
+#     # 计算宏平均PR曲线和平均精度
+#     precision["macro"] = np.mean([precision[i] for i in range(conf.n_classes)], axis=0)
+#     recall["macro"] = np.mean([recall[i] for i in range(conf.n_classes)], axis=0)
+#     average_precision["macro"] = np.mean([average_precision[i] for i in range(conf.n_classes)])
+#
+#     # 绘制微平均PR曲线
+#     plt.figure(figsize=(10, 7))
+#     plt.plot(recall["micro"], precision["micro"], color='gold', lw=2,
+#              label='micro-average Precision-recall curve (area = {0:0.2f})'
+#                    ''.format(average_precision["micro"]))
+#
+#     # 绘制宏平均PR曲线
+#     plt.plot(recall["macro"], precision["macro"], color='navy', lw=2,
+#              label='macro-average Precision-recall curve (area = {0:0.2f})'
+#                    ''.format(average_precision["macro"]))
+#
+#     # 绘制每个类别的PR曲线
+#     colors = plt.cm.tab20(np.linspace(0, 1, conf.n_classes))
+#     for i, color in zip(range(conf.n_classes), colors):
+#         plt.plot(recall[i], precision[i], color=color, lw=2,
+#                  label='Precision-recall curve of class {0} (area = {1:0.2f})'
+#                        ''.format(i, average_precision[i]))
+#
+#     plt.xlabel('Recall')
+#     plt.ylabel('Precision')
+#     plt.title('Extension of Precision-Recall curve to multi-class')
+#     plt.legend(loc="best")
+#     plt.show()
+#     pass
+
+
+def calculate_similarity(outputs, targets, threshold):
+    FN, FP, TN, TP = 0, 0, 0, 0
+    for output, target in zip(outputs, targets):
+        # print("output >> {} , target >> {}".format(output, target))
+        if output != target:
+            if target == 0:
+                FP += 1
+            elif target == 1:
+                FN += 1
+        else:
+            if target == 0:
+                TN += 1
+            elif target == 1:
+                TP += 1
+    if TP == 0:
+        prec, recall = 0, 0
+    else:
+        prec = TP / (TP + FP)
+        recall = TP / (TP + FN)
+    if TN == 0:
+        tn_prec, tn_recall = 0, 0
+    else:
+        tn_prec = TN / (TN + FN)
+        tn_recall = TN / (TN + FP)
+        # print("TP>>{}, FP>>{}, TN>>{}, FN>>{}".format(TP, FP, TN, FN))
+    if threshold == 0.1:
+        print("TP>>{}, FP>>{}, TN>>{}, FN>>{}".format(TP, FP, TN, FN))
+    return prec, recall, tn_prec, tn_recall
+
+
+def show_pr(prec, recall, tn_prec, tn_recall, thres, title_name):
+    x = thres
+    plt.figure(figsize=(10, 6))
+    plt.plot(x, recall, color='red', label='recall:TP/TPFN')
+    plt.plot(x, tn_recall, color='black', label='recall_TN:TN/TNFP')
+    plt.plot(x, prec, color='blue', label='PrecisePos:TP/TPFN')
+    plt.plot(x, tn_prec, color='green', label='PreciseNeg:TN/TNFP')
+    plt.legend()
+    plt.xlabel('threshold')
+    # if self.title_name is not None:
+    plt.title(title_name)
+    plt.grid(True, linestyle='--', alpha=0.5)
+    plt.savefig(os.sep.join(['../pr_test', 'yolo_pr.png']))
+    plt.show()
+    plt.close()
+
+
+def write_results_to_file(title_name, thres, recall, recall_TN, PrecisePos, PreciseNeg):
+    file_path = os.sep.join(['../pr_test', title_name + '.txt'])
+    with open(file_path, 'w') as file:
+        file.write("threshold,   recall,   recall_TN,   PrecisePos,   PreciseNeg\n")
+        for thre, rec, rec_tn, prec_pos, prec_neg in zip(thres, recall, recall_TN, PrecisePos, PreciseNeg):
+            file.write(
+                f"thre>>{thre:.2f},  recall>>{rec:.4f},  precise_pos>>{prec_pos:.4f},  recall_tn>>{rec_tn:.4f},  precise_neg>>{prec_neg:4f}\n")
+
+
+def show_to_pr(y_true, y_prob, type):  # 二分类
+    thres = [i * 0.01 for i in range(101)]
+    title_name = 'yolo_pr'
+    recall, recall_TN, PrecisePos, PreciseNeg = [], [], [], []
+    for threshold in thres:
+        y_scores_adjusted = []
+        if type == 0:
+            y_scores_adjusted = np.where(y_prob < threshold, 0, 1)
+        elif type == 1:
+            for yp in y_prob:
+                if yp != 0:
+                    yp = np.array(yp)
+                    yp_num = np.sum(np.array(yp) > threshold)
+                    if yp_num / len(yp) > 0.1:
+                        y_scores_adjusted.append(1)
+                    else:
+                        y_scores_adjusted.append(0)
+                else:
+                    y_scores_adjusted.append(0)
+        prec, pos_recall, tn_prec, tn_recall = calculate_similarity(y_scores_adjusted, y_true, threshold)
+        recall.append(pos_recall)
+        recall_TN.append(tn_recall)
+        PrecisePos.append(prec)
+        PreciseNeg.append(tn_prec)
+        # print(" prec>>{}  recall>>{}   tn_prec>>{}  tn_recall>>{} threshold>>{}\n".format(prec, pos_recall, tn_prec,
+        #                                                                                   tn_recall, threshold))
+    show_pr(PrecisePos, recall, PreciseNeg, recall_TN, thres, title_name)
+    write_results_to_file(title_name, thres, recall, recall_TN, PrecisePos, PreciseNeg)
+
+
+if __name__ == '__main__':
+    pass