Update app.py

app.py

@@ -3,38 +3,49 @@ import numpy as np
 import pickle
 import streamlit.components.v1 as components
 from sklearn.feature_extraction.text import CountVectorizer
+from sklearn.model_selection import train_test_split
+import pandas as pd
+# Separate target and feature column in X and y variable
+df = pd.read_json('sarcasm.json',lines = True)
+# X will be the features
+X = np.array(df["headline"])
 
+# y will be the target variable
+y = np.array(df["is_sarcastic"])
+cv = CountVectorizer()
 # Load the pickled model
-def load_model():
-    return pickle.load(open('Sarcasm_Detection_BernoulliNB.pkl', 'rb')) #Load model from notebook
+
+X = cv.fit_transform(X)
+X_train, X_test, y_train, y_test = train_test_split(X, y,
+                                                test_size=0.33,
+                                                random_state=42)
 
 # Function for model prediction
-def model_prediction(model, features):
-    features = model.transform([features])
-    predicted = str(model.predict([features]))
-    return predicted
+def model_prediction(features):
+    features = cv.transform([features]).toarray()
+    pickled_model = pickle.load(open('Sarcasm_Detection_BernoulliNB.pkl', 'rb'))
+    Message = str(list(pickled_model.predict(features)))
+
+    return Message
 
 def app_design():
     # Add input fields for High, Open, and Low values
-    image = '32.png' # Load image 
-    st.image(image, use_column_width=True)
+    # image = '21.png' # Load image 
+    # st.image(image, use_column_width=True)
     
     st.subheader("Enter the following values:")
 
     text= st.text_input("Enter your message")
     # Create a feature list from the user inputs
-    features = [[text]] # add features according to notebook
-    
-    # Load the model
-    model = load_model()
+    features = text # add features according to notebook
     
     # Make a prediction when the user clicks the "Predict" button
     if st.button('Predict Sarcasm'):  
-        predicted_value = model_prediction(model, features)
-        if predicted_value == 'Sarcasm':
-            st.success(f"Your message contains Sarcasm") 
-        else:
-              st.success(f"Your message doesnot contains Sarcasm")       
+        predicted_value = model_prediction(features)
+        if predicted_value == "['Sarcasm']":
+            st.success("Your message contains Sarcasm") 
+        elif predicted_value ==  "['Not Sarcasm']":
+              st.success("Your message doesnot contains Sarcasm")