JupyterN

import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import SimpleRNN, Dense

# Data preparation in mmol/L
glucose_values_mmol = np.array([5.0, 4.4, 5.3, 4.2, 3.9, 4.7, 5.3, 4.7, 4.4, 3.9, 4.2])

# Data normalization
glucose_normalized = (glucose_values_mmol - np.min(glucose_values_mmol)) / (np.max(glucose_values_mmol) - np.min(glucose_values_mmol))

# Prepare training data
X = glucose_normalized[:-1]
y = glucose_normalized[1:]

# Create the model
model = Sequential([
    SimpleRNN(units=1, input_shape=(1, 1), activation='linear'),
    Dense(units=1, activation='linear')
])

# Compile the model
model.compile(optimizer='adam', loss='mean_squared_error')

# Reshape the data for training
X = X.reshape(-1, 1, 1)
y = y.reshape(-1, 1)

# Train the model
model.fit(X, y, epochs=100)

# Predict based on previous values
X_test = X  # Using the same data for prediction
y_pred_normalized = model.predict(X_test)
y_pred = y_pred_normalized * (np.max(glucose_values_mmol) - np.min(glucose_values_mmol)) + np.min(glucose_values_mmol)

# Display the results
x_values = np.arange(1, len(glucose_values_mmol))
x_values_pred = np.arange(2, len(glucose_values_mmol) + 1)

plt.plot(x_values, glucose_values_mmol[:-1], marker='o', label='Actual values')
plt.plot(x_values_pred, y_pred, marker='o', label='Prediction')
plt.legend()
plt.show()