104 lines
4.6 KiB
Python
104 lines
4.6 KiB
Python
from experiments.experiment import experiments
|
||
from concurrent.futures import ProcessPoolExecutor
|
||
from loadData_and_MakeImageGenerator.load_and_ImageGenerator import Load_ImageGenerator
|
||
from Read_and_process_image.ReadAndProcess import Read_image_and_Process_image
|
||
import tensorflow as tf
|
||
import numpy as np
|
||
from sklearn.metrics import confusion_matrix, accuracy_score
|
||
from draw_tools.draw import draw_heatmap
|
||
from Load_process.LoadData import Loding_Data_Root
|
||
import os
|
||
import seaborn as sns
|
||
import datetime
|
||
from Load_process.file_processing import judge_file_exist, make_dir, make_save_root
|
||
from matplotlib import pyplot as plt
|
||
from Load_process.Load_Indepentend import Load_Indepentend_Data
|
||
|
||
def draw(matrix, model_name, index):
|
||
# Using Seaborn heatmap to create the plot
|
||
|
||
fx = sns.heatmap(matrix, annot=True, cmap='turbo')
|
||
# labels the title and x, y axis of plot
|
||
fx.set_title('Plotting Confusion Matrix using Seaborn\n\n')
|
||
fx.set_xlabel('Predicted Values')
|
||
fx.set_ylabel('answer Values ')
|
||
# labels the boxes
|
||
fx.xaxis.set_ticklabels(['False','True'])
|
||
fx.yaxis.set_ticklabels(['False','True'])
|
||
|
||
model_dir = '../../Model_Confusion_matrix/model_matrix_image ( ' + str(datetime.date.today()) + " )/" + model_name
|
||
if not judge_file_exist(model_dir):
|
||
make_dir(model_dir)
|
||
modelfiles = make_save_root(str(model_name) + "-" + str(index) + ".png", model_dir)
|
||
plt.savefig(modelfiles)
|
||
plt.close("all") # 關閉圖表
|
||
|
||
if __name__ == "__main__":
|
||
with ProcessPoolExecutor() as executor: ## 默认为1
|
||
print('TensorFlow version:', tf.__version__)
|
||
physical_devices = tf.config.experimental.list_physical_devices('GPU')
|
||
print(physical_devices)
|
||
assert len(physical_devices) > 0, "Not enough GPU hardware devices available"
|
||
tf.config.experimental.set_memory_growth(physical_devices[0], True)
|
||
os.environ["CUDA_VISIBLE_DEVICES"]='0'
|
||
|
||
load = Loding_Data_Root()
|
||
experiment = experiments()
|
||
image_processing = Read_image_and_Process_image()
|
||
cut_image = Load_Indepentend_Data()
|
||
|
||
model = experiment.construct_model()
|
||
model_dir = '../../best_model( 2023-10-17 )-2.h5' # 這是一個儲存模型權重的路徑，每一個模型都有一個自己權重儲存的檔
|
||
if os.path.exists(model_dir): # 如果這個檔案存在
|
||
model.load_weights(model_dir) # 將模型權重讀出來
|
||
print("讀出權重\n")
|
||
|
||
for times in range(5):
|
||
name = ["BP", "PF", "PV", "Chickenpox", "Monkeypox", "Normal", "Another"]
|
||
cut_image.process_main() # 呼叫處理test Data與Validation Data的function
|
||
test, test_label = cut_image.test, cut_image.test_label
|
||
|
||
total_data = [[], [], [], [], [], [], []]
|
||
total_labels = [[], [], [], [], [], [], []]
|
||
start = 0
|
||
end = 22
|
||
for k in range(7):
|
||
for i in range(start, end):
|
||
total_data[k].append(test[i])
|
||
total_labels[k].append(test_label[i])
|
||
|
||
total_data[k], total_labels[k] = image_processing.image_data_processing(total_data[k], total_labels[k])
|
||
total_data[k] = image_processing.normalization(total_data[k])
|
||
|
||
start = end
|
||
end += 22
|
||
|
||
|
||
j = 0
|
||
for total_label in range(7):
|
||
result = model.predict(total_data[j]) # 利用predict function來預測結果
|
||
result = np.argmax(result, axis = 1) # 將預測出來的結果從one-hot encoding轉成label-encoding
|
||
y_test = np.argmax(total_labels[j], axis = 1)
|
||
|
||
print(name[j] + str(result), "\n")
|
||
|
||
y_pre = []
|
||
for i in range(len(result)):
|
||
if result[i] != j:
|
||
result[i] = 0
|
||
else:
|
||
result[i] = 1
|
||
|
||
y_test[i] = 1
|
||
|
||
matrix = confusion_matrix(y_test, result, labels = [0, 1]) # 丟入confusion matrix的function中，以形成混淆矩陣
|
||
draw(matrix, name[j], times) # 呼叫畫出confusion matrix的function
|
||
tn, fp, fn, tp = matrix.ravel()
|
||
|
||
accuracy = (tn + tp) / (tn + fp + fn + tp)
|
||
print(name[j] + " 權重為: ", accuracy)
|
||
|
||
experiment.record_everyTime_test_result(0, accuracy, 0, 0, 0, 0, times, name[j])
|
||
|
||
j += 1
|