Added file for the oral presentation when we will have to test

Main/CC3TestFile.py

+from Pipeline.ImageFolderManager import buildSampleFromPath, fetch_images_to_dict
+from Pipeline.Model import get_SVC_model_with_best_parameters
+from Pipeline.Evaluation import cross_validation_on_model
+from Pipeline.AlgoTraining import fit
+from Pipeline.PredictionFile import computePredictionFile
+from joblib import Memory
+
+PATH_SEA = "../Init/Mer"  # PATH TO THE IMAGES CONTAINING THE SEA
+PATH_WITHOUT_SEA = "../Init/Ailleurs"  # PATH TO THE IMAGE WITHOUT THE SEA
+
+memory = Memory("cached_memory")
+
+@memory.cache()
+def classifier(model):
+    S = buildSampleFromPath(PATH_SEA, PATH_WITHOUT_SEA)
+    return fit(S, model), S
+
+
+algo, S = classifier(get_SVC_model_with_best_parameters())
+print("loaded trained algorithm")
+path = "../Init/Mer" #TODO METTRE LE NOM DU DOSSIER
+result_file_name = "testFile.txt" #TODO METTRE LE NOM DU FICHIER
+
+print("loading test images")
+images_to_test = fetch_images_to_dict(path)
+print(f"predicting and computing result file named {result_file_name}")
+computePredictionFile(algo, images_to_test, result_file_name, S)

Pipeline/AlgoTraining.py

@@ -24,6 +24,17 @@ def fit_algorithm(S, algo):
 
     return algo, S_test, y_test, S_train, y_train
 
+def fit(S, algo):
+    """
+    Same function as 'fit_algorithm' however, you only get the classifier trained on the sample S. It's mainly used
+    for simplicity in the usage.
+    :param S: the sample on which we train
+    :param algo: the algo to fit the data on
+    :return: the fitted algorithm
+    """
+    algo, S_test, y_test, S_train, y_train = fit_algorithm(S, algo)
+    return algo
+
 
 def predict(S, model, list_dict=True):
     """

Pipeline/Evaluation.py

@@ -5,6 +5,7 @@ from sklearn.metrics import accuracy_score
 from Pipeline.ImageDictionaryManager import extract_relevant_data
 from Pipeline.AlgoTraining import fit_algorithm
 
+
 def computeError(S):
     """
     Compute the empirical error of the model on the given sample.
@@ -66,3 +67,5 @@ def cross_validation_on_model(model, S):
     allowing us to have a way more precise score function.
     """
     print("Taux de réussite en cross-validation: ", get_cross_val_score(model, S), "%")
+
+

Pipeline/PredictionFile.py

+import os
+
+from Pipeline.ImageFolderManager import buildSampleFromPath
+from Pipeline.AlgoTraining import fit_algorithm, predict
+from Pipeline.Evaluation import computeError
+
+def computePredictionFile(classifier, images_test=None, filename=None, S=None):
+    """
+    Compute a file with the predictions of the model.
+    :param classifier: the classifier to use.
+    :return: A file .txt with the name of the team, the classifier used, the description methods of images,
+    the prediction on each image, the empirical error, the real error and the score.
+    """
+    file = open(filename if filename is not None else "syntax_error.txt" , "w+")
+    file.write("# T. Bertochio, N. Efremon, P. Vigeolas-Choury (Equipe Syntax Error) \n")
+    file.write("# Nom de l’algorithme d’apprentissage utilisé: SVC  \n")
+    file.write("# Valeurs des hyper-paramètres principaux de l’algorithme: (optimisé GridSearch) \n")
+    file.write("# Concaténation: histogramme de couleurs, coefficients GLCM, filtre de Gabor \n")
+
+    S = S if S is not None else buildSampleFromPath("../Init/Mer", "../Init/Ailleurs")
+    classifier, S_test, y_test, S_train, y_train = fit_algorithm(S, classifier)
+    predict(S, classifier)
+    predict(images_test, classifier)
+
+    images = S if images_test is None else images_test
+    for image in images:
+        image_name = os.path.split(image["name_path"])[1]
+        if image["y_predicted_class"] == 1:
+            image_predicted_class = "+1"
+        else:
+            image_predicted_class = "-1"
+        file.write(image_name+" "+image_predicted_class+"\n")
+
+    file.write("# EE = "+ str(computeError(S_train))+"\n")
+    file.write("# ER = "+ str(computeError(S_test))+"\n")
+    file.close()
\ No newline at end of file

test_for_me.py

@@ -8,8 +8,9 @@ path = './Init/Mer'
 images_path = os.listdir(path)
 full_path = os.path.join(path, images_path[0])
 image = Image.open(full_path)
-image = TP.resizeImage(image.convert("RGB"), 224, 224)
+image.save("image.png")
+image = image.convert("L")
 image.show()
-image.save("test_no_rotate.png")
+image.save("image_gray.png")
 image_r = image.rotate(90)
 image_r.save("test_rotate.png")