From 77eb99cbf58fd4212c75fd412cf7330bb1c3694e Mon Sep 17 00:00:00 2001
From: fferreira <fferreira@isptgroup.com>
Date: Mon, 1 Sep 2025 21:43:35 +0100
Subject: [PATCH] first commit

---
 Model.py   | 31 +++++++++++++++++++++++++++++++
 Predict.py | 14 ++++++++++++++
 graph.py   | 42 ++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 87 insertions(+)
 create mode 100644 Model.py
 create mode 100644 Predict.py
 create mode 100644 graph.py

diff --git a/Model.py b/Model.py
new file mode 100644
index 0000000..b8817bc
--- /dev/null
+++ b/Model.py
@@ -0,0 +1,31 @@
+import pandas as pd
+import numpy as np
+
+# Make numpy values easier to read.
+np.set_printoptions(precision=3, suppress=True)
+
+import tensorflow as tf
+from tensorflow.keras import layers
+
+ecvalve = pd.read_csv(
+    "/home/fferreira/Documents/teste_ecvalve.csv",
+    names=[ "frequency", "state", "pressure", "capacity"])
+print("Data shape:", ecvalve.shape)
+print("Columns:", ecvalve.columns)
+
+ecvalve.head()
+ecvalve_features = ecvalve.copy()
+ecvalve_labels = ecvalve_features.pop('state')
+
+ecvalve_features = np.array(ecvalve_features)
+
+ecvalve_model = tf.keras.Sequential([
+    layers.Dense(64, activation='relu', input_shape=(ecvalve_features.shape[1],)),
+    layers.Dense(32, activation='relu'),  # Add another layer
+    layers.Dense(1, activation='sigmoid')  # Sigmoid for binary classification
+])
+ecvalve_model.compile(loss=tf.keras.losses.BinaryCrossentropy(), optimizer=tf.keras.optimizers.Adam(), metrics=['accuracy'])
+
+ecvalve_model.fit(ecvalve_features, ecvalve_labels, epochs=20)
+
+ecvalve_model.save('EcValve_1.h5')
\ No newline at end of file
diff --git a/Predict.py b/Predict.py
new file mode 100644
index 0000000..666ebd7
--- /dev/null
+++ b/Predict.py
@@ -0,0 +1,14 @@
+import tensorflow as tf
+import numpy as np
+
+import tensorflow as tf
+import numpy as np
+
+model = tf.keras.models.load_model('EcValve_1.h5')
+
+input_data = np.array([[-0.029184604, 7, 60]], dtype=np.float32)  # Shape: (1, 3)
+
+prediction = model.predict(input_data)
+
+print("Prediction (probability of leak):", prediction)
+print("Leak detected" if prediction[0] > 0.5 else "No leak")
\ No newline at end of file
diff --git a/graph.py b/graph.py
new file mode 100644
index 0000000..eac7c65
--- /dev/null
+++ b/graph.py
@@ -0,0 +1,42 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+
+# Set seaborn style for better visuals
+sns.set(style="whitegrid")
+
+# Load CSV
+ecvalve = pd.read_csv(
+    "/home/fferreira/Documents/teste_ecvalve.csv",
+    names=["frequency", "state", "pressure", "capacity"])
+print("Data shape:", ecvalve.shape)
+print("Label distribution:", ecvalve['state'].value_counts())
+
+# Group by frequency and get unique states
+freq_states = ecvalve.groupby('frequency')['state'].unique().reset_index()
+freq_states['category'] = freq_states['state'].apply(
+    lambda x: '0' if len(x) == 1 and x[0] == 0 else '1' if len(x) == 1 and x[0] == 1 else 'Both'
+)
+
+print("\nFrequencies and their state categories:")
+print(freq_states[['frequency', 'category']].sort_values('frequency'))
+
+# Create scatter plot
+plt.figure(figsize=(10, 6))
+sns.scatterplot(
+    x='frequency', 
+    y='category', 
+    hue='category', 
+    style='category', 
+    data=freq_states, 
+    palette={'0': '#36A2EB', '1': '#FF6384', 'Both': '#FFCE56'}, 
+    s=100
+)
+plt.title('Frequencies by State Category (0 = No Leak, 1 = Leak, Both = Mixed)')
+plt.xlabel('Frequency (Hz)')
+plt.ylabel('State Category')
+plt.savefig('/home/fferreira/Documents/HarvardX/teste/frequency_state_categories.png')
+plt.close()
+
+print("Graph saved: /home/fferreira/Documents/HarvardX/teste/frequency_state_categories.png")
\ No newline at end of file