Initial Commit

app.py

@@ -0,0 +1,44 @@
+import gradio as gr
+import numpy as np
+import pickle
+from tensorflow.keras.models import load_model
+
+# Load the saved model and scaler
+model = load_model('Best_GRU_solar_insolation_model.keras')
+with open('scaler.pkl', 'rb') as f:
+    scaler = pickle.load(f)
+
+# Preprocessing function for Gradio input (scaling + sin/cos transformations)
+def preprocess_input(year, month, day, hour):
+    # Sin/Cos transformations
+    day_sin = np.sin(2 * np.pi * day / 31)
+    day_cos = np.cos(2 * np.pi * day / 31)
+    month_sin = np.sin(2 * np.pi * month / 12)
+    month_cos = np.cos(2 * np.pi * month / 12)
+    hour_sin = np.sin(2 * np.pi * hour / 24)
+    hour_cos = np.cos(2 * np.pi * hour / 24)
+    
+    # Combine the features
+    input_data = np.array([year, day_sin, day_cos, month_sin, month_cos, hour_sin, hour_cos]).reshape(1, -1)
+    
+    # Scale the input using the saved scaler
+    input_scaled = scaler.transform(input_data)
+    
+    # Reshape the scaled input for the model
+    return np.expand_dims(input_scaled, axis=-1)
+
+# Prediction function for Gradio
+def predict_solar_insolation(year, month, day, hour):
+    input_data = preprocess_input(year, month, day, hour)
+    prediction = model.predict(input_data)
+    return round(float(prediction[0][0]), 3) if float(prediction[0][0])>10 else 0
+
+# Gradio client
+app = gr.app(fn=predict_solar_insolation,
+                         inputs=[gr.Number(label="Year"),
+                                 gr.Number(label="Month"),
+                                 gr.Number(label="Day"),
+                                 gr.Number(label="Hour")],
+                         outputs="number")
+
+app.launch()

requirements.txt

@@ -0,0 +1,6 @@
+gradio
+numpy
+pandas
+scikit-learn
+tensorflow
+matplotlib

scaler.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:937b91f3028ec6b61b4131e5877e7218eafccc4091e895a236542edfcc6d4bf8
+size 923