-
Notifications
You must be signed in to change notification settings - Fork 2
/
train.py
65 lines (43 loc) · 1.95 KB
/
train.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
import utils
import pandas as pd
import random
import numpy as np
from sklearn.model_selection import train_test_split
import model as mdl
data = pd.read_csv('Dx_map.csv')
df = utils.create_dataframes('training')
srce_files_df = ['cpsc_2018_df', 'cpsc_2018_extra_df', 'georgia_df', 'ptb_df', 'ptb-xl_df', 'st_petersburg_incart_df']
srce_files = ['cpsc_2018', 'cpsc_2018_extra', 'georgia', 'ptb', 'ptb-xl', 'st_petersburg_incart']
#================================================================================================================
X,lengths = utils.create_y_array(srce_files)
Y = utils.create_y_array(df,data,srce_files_df)
#================================================================================================================
# Removing the outliers / Unwanted data
new_sizes = []
for i in range(len(lengths)):
if(lengths[i] < 1000 or lengths[i] > 5000):
Y[i] = 0
X[i] = 0
else:
new_sizes.append(lengths[i])
# Modifying the arrays after removing unwanted values
X = [item for item in X if type(item) != int]
Y = [item for item in Y if type(item) != int]
# Adding noice to the data to make it 2617 points long
X = utils.equalizing_wave_array(X)
# Convering the list of arrays to numpy arrays
for i in range(len(X)):
X[i] = np.array(X[i])
for i in range(len(Y)):
Y[i] = np.array(Y[i])
# Splitting the data into train and test
X_train, X_test, y_train, y_test = train_test_split(np.array(X), np.array(Y), test_size=0.1, random_state=42)
# Getting the input shape and number of classes(output)
input_shape = (X_train.shape[1], X_train.shape[2]) # Shape: (sequence_length, num_leads)
num_classes = y_train.shape[1] # Number of anomaly classes
# Creating the model
resnet_model = mdl.ResNet_model(input_shape,num_classes)
# Training the model
trained_model,accuracy_results_loss_results = mdl.model_train(X_train,y_train,resnet_model,5,15)
# Saving the model
trained_model.save('CardioScanPro_resnet_model.h5')