import run
import sys
import os
import numpy as np
import matplotlib.pyplot as plt
def run_experiment1(data_path):
os.makedirs('exp1_results', exist_ok=True)
accuracies1 = []
accuracies2 = []
accuracies3 = []
accuracies4 = []
accuracies5 = []
for num_subjects in range(20, 150, 20):
print(f"Training for {num_subjects} subjects with III+V3+V5")
accuracy, total_train_time, total_eval_time = run.run(NUM_SEGMENTS=500,
NUM_SECONDS=2,
NUM_BATCH=16,
LEADS=[0,1,2],
NUM_EPOCHS=200,
DATA_PATH = data_path,
FS=128,
NUM_PATIENTS=num_subjects)
accuracies1.append(accuracy)
with open('exp1_results/result-III-V3-V5.txt', 'a') as f:
f.write(f"With number subjects: {num_subjects}:\n")
f.write(f"Accuracy: {accuracy}, Time train: {total_train_time}, Time eval: {total_eval_time}\n\n")
for num_subjects in range(20, 150, 20):
print(f"Training for {num_subjects} subjects with III+V3")
accuracy, total_train_time, total_eval_time = run.run(NUM_SEGMENTS=500,
NUM_SECONDS=2,
NUM_BATCH=16,
LEADS=[0,1],
NUM_EPOCHS=200,
DATA_PATH = data_path,
FS=128,
NUM_PATIENTS=num_subjects)
accuracies2.append(accuracy)
with open('exp1_results/result-III-V3.txt', 'a') as f:
f.write(f"With number subjects: {num_subjects}:\n")
f.write(f"Accuracy: {accuracy}, Time train: {total_train_time}, Time eval: {total_eval_time}\n\n")
for num_subjects in range(20, 150, 20):
print(f"Training for {num_subjects} subjects with V5")
accuracy, total_train_time, total_eval_time = run.run(NUM_SEGMENTS=500,
NUM_SECONDS=2,
NUM_BATCH=16,
LEADS=[2],
NUM_EPOCHS=200,
DATA_PATH = data_path,
FS=128,
NUM_PATIENTS=num_subjects)
accuracies3.append(accuracy)
with open('exp1_results/result-V5.txt', 'a') as f:
f.write(f"With number subjects: {num_subjects}:\n")
f.write(f"Accuracy: {accuracy}, Time train: {total_train_time}, Time eval: {total_eval_time}\n\n")
for num_subjects in range(20, 150, 20):
print(f"Training for {num_subjects} subjects with V3")
accuracy, total_train_time, total_eval_time = run.run(NUM_SEGMENTS=500,
NUM_SECONDS=2,
NUM_BATCH=16,
LEADS=[1],
NUM_EPOCHS=200,
DATA_PATH = data_path,
FS=128,
NUM_PATIENTS=num_subjects)
accuracies4.append(accuracy)
with open('exp1_results/result-V3.txt', 'a') as f:
f.write(f"With number subjects: {num_subjects}:\n")
f.write(f"Accuracy: {accuracy}, Time train: {total_train_time}, Time eval: {total_eval_time}\n\n")
for num_subjects in range(20, 150, 20):
print(f"Training for {num_subjects} subjects with III")
accuracy, total_train_time, total_eval_time = run.run(NUM_SEGMENTS=500,
NUM_SECONDS=2,
NUM_BATCH=16,
LEADS=[0],
NUM_EPOCHS=200,
DATA_PATH = data_path,
FS=128,
NUM_PATIENTS=num_subjects)
accuracies5.append(accuracy)
with open('exp1_results/result-III.txt', 'a') as f:
f.write(f"With number subjects: {num_subjects}:\n")
f.write(f"Accuracy: {accuracy}, Time train: {total_train_time}, Time eval: {total_eval_time}\n\n")
#plotting
num_subjects = list(range(20, 150, 20))
plt.figure(figsize=(10, 8), dpi=100)
plt.plot(num_subjects, accuracies1, '#8C2155',marker='.', label='III+V3+V5')
plt.plot(num_subjects, accuracies2, '#F99083',marker='.', label='III+V3')
plt.plot(num_subjects, accuracies3, '#8AA29E',marker='.', label='V5')
plt.plot(num_subjects, accuracies4, '#98CE00',marker='.', label='V3')
plt.plot(num_subjects, accuracies5, '#FFC857',marker='.', label='III')
plt.xlabel('Number of subjects', fontsize=16)
plt.ylabel('Accuracy (%)', fontsize=16)
plt.tick_params(axis='both', which='major', labelsize=12)
plt.legend()
plt.xlim((np.min(num_subjects),np.max(num_subjects)))
plt.grid(True, 'both')
plt.savefig('exp1_results/fig_exp1.png')
plt.show()
plt.close()
if __name__ == "__main__":
if len(sys.argv) != 2:
print("Usage: python run_exp_1.py <data_path>")
sys.exit(1)
data_path = sys.argv[1]
run_experiment1(data_path)
Datasets
Models
