src/experiments/decision_tree/decision_tree.py

@@ -1,14 +1,14 @@
-# models
+from enum import Enum
 from sklearn import tree
+from sklearn import metrics
 from sklearn import preprocessing
 from sklearn import neighbors
 from sklearn import ensemble
 from sklearn import svm
-
-# other
-from enum import Enum
+from matplotlib import pyplot as plt
+import pandas as pd
 import numpy as np
-import time
+import random
 import csv
 import plots
 
@@ -49,68 +49,58 @@ with open(PATH, 'r') as file:
     normalized = preprocessing.normalize(data, axis=0, norm='max')
     norm = list(normalized.tolist())
 
-steps = np.linspace(1e-4, 1, 20, dtype=np.float64)
+steps = np.linspace(0.1, 1.0, 10, dtype=np.float64)
 
-print("Step \t seconds/step")
 for step in steps:
     actual = []
     predicted = []
-    time_start = time.time()
 
-    for j in range(3):
-        for i in range(len(norm)):
-            temp_data = norm.pop(i)
-            temp_label = labels.pop(i)
+    for i in range(len(norm)):
+        temp_data = norm.pop(i)
+        temp_label = labels.pop(i)
 
-            # model = tree.DecisionTreeClassifier(
-            #     class_weight=None,
-            #     min_samples_leaf=2,
-            #     max_depth=None, # < 5 is worse, None good too
-            #     random_state=False, # No change
-            #     criterion='gini', # MCC + 0.1
-            #     splitter='best',
-            #     ccp_alpha=0 # Pruning: Keep this 0
-            # )
-            # model = ensemble.RandomForestClassifier(
-            #     n_estimators=20, # higher is better, but slower (def: 100)
-            #     criterion='gini', # gini best
-            # )
-            # model = ensemble.ExtraTreesClassifier(
-            #     n_estimators=step # higher is better, but slower (def: 100)
-            # )
-            # model = neighbors.KNeighborsClassifier(
-            #     algorithm='auto',
-            #     leaf_size=2,
-            #     n_neighbors=step,
-            # )
-            # model = ensemble.BaggingClassifier(
-            #     n_estimators=5,
-            #     max_samples=.5,
-            #     max_features=.5,
-            #     bootstrap=False
-            # )
-            # model = svm.SVC(
-            #     C = 0.8,
-            #     kernel = "poly",
-            #     degree = 5,
-            #     coef0 = 6,
-            #     probability = False,
-            #     break_ties=True,
-            #     decision_function_shape = 'ovr'
-            # )
-            model = model.fit(norm, labels)
-            result = model.predict([temp_data])
+        # model = tree.DecisionTreeClassifier(
+        #     class_weight=None,
+        #     min_samples_leaf=2,
+        #     max_depth=None, # < 5 is worse, None good too
+        #     random_state=False, # No change
+        #     criterion='gini', # MCC + 0.1
+        #     splitter='best',
+        #     ccp_alpha=0 # Pruning: Keep this 0
+        # )
+        # model = ensemble.RandomForestClassifier(
+        #     n_estimators=20, # higher is better, but slower (def: 100)
+        #     criterion='gini', # gini best
+        # )
+        # model = ensemble.ExtraTreesClassifier(
+        #     n_estimators=150 # higher is better, but slower (def: 100)
+        # )
+        # model = neighbors.KNeighborsClassifier(
+        #     algorithm='auto',
+        #     leaf_size=2,
+        #     n_neighbors=step,
+        # )
+        model = ensemble.BaggingClassifier(
+            n_estimators=5,
+            max_samples=.5,
+            max_features=.5,
+            bootstrap=False
+        )
+        # model = svm.SVC(decision_function_shape='ovr'
+        # )
+        model = model.fit(norm, labels)
+        result = model.predict([temp_data])
 
-            norm.append(temp_data)
-            labels.append(temp_label)
+        norm.append(temp_data)
+        labels.append(temp_label)
 
-            actual.append(temp_label)
-            predicted.append(result[0])
+        actual.append(temp_label)
+        predicted.append(result[0])
 
     actual_list.append(actual)
     predicted_list.append(predicted)
 
-    print("%.4f"%step, "\t", "%.2f"%(time.time()-time_start))
+    print(step)
 
 plots.plotMetrics(actual_list, predicted_list)
 plots.plotConfusion(actual_list[0], predicted_list[0])

src/experiments/template_extraction/script.py

@@ -94,7 +94,6 @@ detector = cv2.aruco.ArucoDetector(dictionary, detector_params)
 
 images_converted = 0
 images_skipped = 0
-names_skipped = []
 
 ### IMAGE CONVERSIE ###
 for folder in os.listdir(input_directory):
@@ -148,12 +147,16 @@ for folder in os.listdir(input_directory):
         if VERBOSE:
             print("IDs detected:\n", ids)
 
-        if ids is None or len(ids) != 4:
+        if ids is None:
             print("Skipping: ", filename)
             print("=============================================")
-            names_skipped.append(filename)
             images_skipped += 1
             continue
+        if len(ids) != 4:
+            print("Skipping: ", filename)
+            print("=============================================")
+            images_skipped += 1
+            continue 
 
         if VERBOSE:
             print("%d markers gedetecteerd" %len(ids))
@@ -253,12 +256,5 @@ for folder in os.listdir(input_directory):
 if VERBOSE:
     print("%d van de %d succesvol"
           %(images_converted, (images_converted+images_skipped)))
-    
-if images_skipped != 0:
-    print("")
-    with open(os.path.join(input_directory, "skipped.txt"), 'w') as file:
-        for name in names_skipped:
-            file.write(name)
-            file.write("\n")
 
 cv2.destroyAllWindows()