load two layer of cnn

_mian_tensorflow_keras_.py

@@ -1,92 +0,0 @@
-import numpy as np
-import pywt
-import wfdb
-import tensorflow as tf
-
-ECGDATAPATH= "mit-bih-arrhythmia-database-1.0.0/"
-
-
-def denoise(date):
-    # Wavelet db5 level9
-    coeffs = pywt.wavedec(data=date, wavelet='db5', level=9)
-    cA9, cD9, cD8, cD7, cD6, cD5, cD4, cD3, cD2, cD1 = coeffs
-    # noice del
-    threshold = (np.median(np.abs(cD1)) / 0.6745) * (np.sqrt(2 * np.log(len(cD1))))
-    # take cD 1,2 turn 0 to denoise
-    cD1.fill(0)
-    cD2.fill(0)
-    for i in range(1, len(coeffs) - 2):
-        coeffs[i] = pywt.threshold(coeffs[i], threshold)
-    # get signal that before wavelet
-    rdata = pywt.waverec(coeffs=coeffs, wavelet='db5')
-    return rdata
-
-
-# get ECG data,date type and denoise
-# this function without return
-# use append take return data to pass data
-def getDataSet(number, X_data, Y_data):
-    ecgClassSet = ['N', 'A', 'V', 'l', 'R']
-    print('get' + number + ' ECG data')
-    record = wfdb.rdrecord(ECGDATAPATH + number, channel_names=['MLII'])
-    data = record.p_signal.flatten()
-    # use def denoise
-    rdata = denoise(date=data)
-
-    # get type of ECG data
-    annotation = wfdb.rdann(ECGDATAPATH + number, 'atr')
-    Rlocation = annotation.sample  # annotation.sample:get R
-    Rclass = annotation.symbol
-
-    # delete stable date near R
-    start = 10
-    end = 5
-    i = start
-    j = len(annotation.symbol) - end
-
-    # get NAVLR type data
-    while i < j:
-        try:
-            lable = ecgClassSet.index(Rclass[i])
-            x_train = rdata[Rlocation[i] - 99:Rlocation[i] + 201]
-            X_data.append(x_train)
-            Y_data.append(lable)
-            i += 1
-        except ValueError:
-            i += 1
-        return
-
-
-# RATIO=The ratio of the training set to the test set，usually is 0.3
-def loadData(RATIO):
-    numberSet = [
-        '100', '101', '103', '105', '106', '107', '108', '109', '111', '112', '113', '114', '115', '116', '117', '119',
-        '121', '122', '123', '124', '200', '201', '202', '203', '205', '208', '210', '212', '213', '214', '215', '217',
-        '219', '220', '221', '222', '223', '228', '230', '231', '232', '233', '234'
-    ]
-    dataSet = []
-    labelSet = []
-    for n in numberSet:
-        getDataSet(n, dataSet, labelSet)
-
-    dataSet = np.array(dataSet).reshape(-1, 300)
-    labelSet = np.array(labelSet).reshape(-1, 1)
-    train_ds = np.hstack((dataSet, labelSet))
-    np.random.shuffle(train_ds)
-
-    X = train_ds[:, :300].reshape(-1, 300, 1)
-    Y = train_ds[:, 300]
-    # cut test and train data from ECG data
-    shuffle_index = np.random.permutation(len(X))
-    test_length = int(RATIO * len(shuffle_index))
-    test_index = shuffle_index[:test_length]
-    train_index = shuffle_index[test_index:]
-    X_test, Y_test = X[test_index:], Y[test_index]
-    X_train, Y_train = X[train_index], Y[train_index]
-    return X_train, Y_test, X_test, Y_train
-
-
-class CNN_model_level_one():
-    def __init__(self):
-        super().__init__()
-        self.conv1 = tf.layers.Conv1d(filter=4)

_mian_tensorflow_twoLayerCNN.py

@@ -0,0 +1,113 @@
+import datetime
+import os
+
+import numpy as np
+import pywt
+import wfdb
+import tensorflow as tf
+from tensorflow import keras
+from utils import loadData, plot_history_tf, plot_heat_map
+
+#from ECG_read_weiyuqing_without_wfdb import ECGDATAPATH
+#PATH of this test
+Project_PATH = "./Number-Of-CNN-Layers/Two/"
+#PICTUREPATH= Project_PATH + "picture/"
+log_dir = Project_PATH + "logs/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
+model_Path_one = Project_PATH  + "model/ecg_model_one_layer.h5"
+#model_Path_one = Project_PATH + "ecg_model_one_layer.h5"
+
+RATIO = 0.3
+RANDOM_SEED = 42
+BATCH_SIZE = 128
+NUM_EPOCHS = 30
+
+
+def CNN_model_level_one():
+    leavlOneModel = tf.keras.models.Sequential([
+
+        # #four CNN layers test
+        # tf.keras.layers.InputLayer(input_shape=(300,)),
+        # # reshape the tensor with shape (batch_size, 300) to (batch_size, 300, 1)
+        # tf.keras.layers.Reshape(target_shape=(300, 1)),
+        # # the first convolution layer, 4 21x1 convolution kernels, output shape (batch_size, 300, 4)
+        # tf.keras.layers.Conv1D(filters=4, kernel_size=21, strides=1, padding='SAME', activation='relu'),
+        # # the first pooling layer, max pooling, pooling size=3 , stride=2, output shape (batch_size, 150, 4)
+        # tf.keras.layers.MaxPool1D(pool_size=3, strides=2, padding='SAME'),
+        # # the second convolution layer, 16 23x1 convolution kernels, output shape (batch_size, 150, 16)
+        # tf.keras.layers.Conv1D(filters=16, kernel_size=23, strides=1, padding='SAME', activation='relu'),
+        # # the second pooling layer, max pooling, pooling size=3, stride=2, output shape (batch_size, 75, 16)
+        # tf.keras.layers.MaxPool1D(pool_size=3, strides=2, padding='SAME'),
+        # # the third convolution layer, 32 25x1 convolution kernels, output shape (batch_size, 75, 32)
+        # tf.keras.layers.Conv1D(filters=32, kernel_size=25, strides=1, padding='SAME', activation='relu'),
+        # # the third pooling layer, average pooling, pooling size=3, stride=2, output shape (batch_size, 38, 32)
+        # tf.keras.layers.AvgPool1D(pool_size=3, strides=2, padding='SAME'),
+        # # the fourth convolution layer, 64 27x1 convolution kernels, output shape (batch_size, 38, 64)
+        # tf.keras.layers.Conv1D(filters=64, kernel_size=27, strides=1, padding='SAME', activation='relu'),
+        # # flatten layer, for the next fully connected layer, output shape (batch_size, 38*64)
+        # tf.keras.layers.Flatten(),
+        # # fully connected layer, 128 nodes, output shape (batch_size, 128)
+        # tf.keras.layers.Dense(128, activation='relu'),
+        # # Dropout layer, dropout rate = 0.2
+        # tf.keras.layers.Dropout(rate=0.2),
+        # # fully connected layer, 5 nodes (number of classes), output shape (batch_size, 5)
+        # tf.keras.layers.Dense(5, activation='softmax')
+
+        #take test for one CNN layers model
+
+        tf.keras.layers.InputLayer(input_shape=(300, 1)),
+
+        # ECG data cant add 0 in edge, will broken data ,take false answer. So we need padding same data in edge.
+
+        # take four 21*1 conv excitation function used RElu
+        tf.keras.layers.Conv1D(filters=4, kernel_size=21, strides=1, padding='same', activation='relu'),
+        # take four 3*1 pool,take strids 2.
+        tf.keras.layers.MaxPool1D(pool_size=3, strides=2, padding='same'),
+
+        tf.keras.layers.Conv1D(filters=4, kernel_size=21, strides=1, padding='same', activation='relu'),
+        tf.keras.layers.MaxPool1D(pool_size=3, strides=2, padding='same'),
+        # tf.keras.layers.Conv1D(filters=4, kernel_size=21, strides=1, padding='same', activation='relu'),
+        # tf.keras.layers.MaxPool1D(pool_size=3, strides=2, padding='same'),
+
+        tf.keras.layers.Flatten(),
+        tf.keras.layers.Dense(128, activation='relu'),
+        tf.keras.layers.Dropout(rate=0.2),
+        tf.keras.layers.Dense(5, activation='softmax')
+    ])
+    return leavlOneModel
+
+
+def main():
+    # get traxin test
+    X_train, X_test, y_train, y_test = loadData(RATIO, RANDOM_SEED)
+
+    # get model or create model
+    if os.path.exists(model_Path_one):
+        print(' get model in h5 file')
+        model = tf.keras.models.load_model(filepath=model_Path_one)
+    else:
+        # create new model(if model unexists)
+        model = CNN_model_level_one()
+        model.compile(optimizer='adam',
+                      loss='sparse_categorical_crossentropy',
+                      metrics=['accuracy'])
+        model.summary()
+        # TB make
+        tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir,
+                                                              histogram_freq=1,
+                                                              )
+        # Training and validation
+        history = model.fit(X_train, y_train, epochs=NUM_EPOCHS,
+                            batch_size=BATCH_SIZE,
+                            validation_data=(X_test, y_test),
+                            callbacks=[tensorboard_callback])
+        # validation_split=RATIO,
+        model.save(filepath=model_Path_one)
+        plot_history_tf(history)
+
+    y_pred = np.argmax(model.predict(X_test), axis=-1)
+    plot_heat_map(y_test, y_pred)
+
+
+if __name__ == '__main__':
+    os.environ["CUDA_VISIBLE_DEVICES"] = "-1"
+    main()

noise_take_ecg.py

@@ -1,31 +0,0 @@
-import numpy as np
-import pywt
-import wfdb
-from matplotlib import pyplot as plt
-
-#file error
-
-#use db5 ,level9 do deviose
-from ECG_read_weiyuqing_without_wfdb import ECGDATAPATH
-
-#from ECG_read_weiyuqing_without_wfdb import PATH
-
-record=wfdb.rdrecord(ECGDATAPATH + 100, channel_names=['MLII'])
-data=record.p_signal.flatten()
-
-coeffs=pywt.wavedec(data=data,wavelet='db5',level=9)
-cA9,cD9, cD8, cD7, cD6, cD5, cD4, cD3, cD2, cD1 = coeffs
-
-threshold=(np.median(np.abs(cD1))/0.6745)*(np.sqrt(2*np.log(len(cD1))))
-cD1.fill(0)
-cD2.fill(0)
-for i in range(1,len(coeffs)-2):
-    coeffs[i]=pywt.threshold(coeffs[i],threshold)
-
-rdata=pywt.waverec(coeffs=coeffs,wavelet='db5')
-
-plt.figure(figsize=(20,4))
-plt.plot(data)
-plt.show()
-plt.plot(rdata)
-plt.show()

requirements.txt

@@ -0,0 +1,11 @@
+matplotlib==3.7.0
+numpy==1.23.5
+opencv_python==4.7.0.72
+Pillow==9.4.0
+PyWavelets==1.4.1
+scikit_learn==1.2.1
+seaborn==0.12.2
+tensorflow==2.10.0
+tensorflow_intel==2.11.0
+torch==2.0.0+cu117
+wfdb==4.1.0

some-test/noise_take_ecg.py

@@ -0,0 +1,31 @@
+import numpy as np
+import pywt
+import wfdb
+from matplotlib import pyplot as plt
+
+# file error
+
+# use db5 ,level9 do deviose
+ECGDATAPATH = './MIT-BIT/'
+
+# from ECG_read_weiyuqing_without_wfdb import PATH
+# record = wfdb.rdrecord(ECGDATAPATH + number, channel_names=['MLII'])
+record = wfdb.rdrecord(ECGDATAPATH + '100', channel_names=['MLII'])
+data = record.p_signal.flatten()
+
+coeffs = pywt.wavedec(data=data, wavelet='db5', level=9)
+cA9, cD9, cD8, cD7, cD6, cD5, cD4, cD3, cD2, cD1 = coeffs
+
+threshold = (np.median(np.abs(cD1)) / 0.6745) * (np.sqrt(2 * np.log(len(cD1))))
+cD1.fill(0)
+cD2.fill(0)
+for i in range(1, len(coeffs) - 2):
+    coeffs[i] = pywt.threshold(coeffs[i], threshold)
+
+rdata = pywt.waverec(coeffs=coeffs, wavelet='db5')
+
+plt.figure(figsize=(20, 4))
+plt.plot(data)
+plt.show()
+plt.plot(rdata)
+plt.show()

some-test/wfdb_use.py

@@ -4,14 +4,14 @@ import numpy as np
 import wfdb
 import matplotlib.pyplot
 
-PATH = './mit-bih-arrhythmia-database-1.0.0/'
+PATH = '../MIT-BIT/'
 
 
 # file error with read_ecg_data(data) parameter data false
 
 def read_ecg_data(data):
     # get the data from the ecg data file and set begin and end points
-    record = wfdb.rdrecord(PATH + data, sampfrom=0, sampto=1500)
+    record = wfdb.rdrecord(PATH + 'data', sampfrom=0, sampto=1500)
     # return type record
     print(type(record))
     print(dir(record))

utils.py

@@ -10,7 +10,7 @@ from sklearn.metrics import confusion_matrix
 from sklearn.model_selection import train_test_split
 
 # from _mian_tensorflow import Project_PATH
-Project_PATH = "./Number-Of-CNN-Layers/One/"
+Project_PATH = "./Number-Of-CNN-Layers/Two/"
 
 PICTUREPATH= Project_PATH +'picture/'