Upload 5 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+OnlineNewsPopularity.csv filter=lfs diff=lfs merge=lfs -text

OnlineNewsPopularity.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b66d9088632308cc27fa35af847650d174a5a50503987c4e511de94a99d1c218
+size 24311769

model_features.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f1ad2cc757d88882b181ad900addcc61c0634bacf2fce7ac7c6af32e3b32aa4
+size 353

observatory_app.py

@@ -0,0 +1,178 @@
+import pandas as pd
+import numpy as np
+from sklearn.linear_model import Ridge
+from sklearn.model_selection import train_test_split
+import joblib
+df = pd.read_csv("OnlineNewsPopularity.csv")
+df.columns = df.columns.str.strip()
+
+df['log_shares'] = np.log1p(df['shares'])
+
+feature_cols = [
+    'n_tokens_content', 'num_imgs', 'global_sentiment_polarity', 
+    'global_subjectivity', 'title_sentiment_polarity', 
+    'data_channel_is_tech', 'weekday_is_monday', 'weekday_is_tuesday', 'weekday_is_wednesday',
+    'weekday_is_thursday', 'weekday_is_friday', 'weekday_is_saturday', 'weekday_is_sunday', "n_tokens_title", "num_videos", "num_keywords", 
+    "num_imgs", "num_hrefs"
+]
+x = df[feature_cols]
+y = df['log_shares']
+
+X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=42)
+model = Ridge(alpha=1.0)
+model.fit(X_train, y_train)
+
+joblib.dump(model, 'popularity_model.pkl')
+joblib.dump(x.columns.tolist(), 'model_features.pkl')
+
+from textblob import TextBlob
+import numpy as np
+import joblib
+
+# Load the trained model and features (Assumes Step 1 was run with the new feature list)
+MODEL = joblib.load('popularity_model.pkl')
+FEATURE_COLUMNS = joblib.load('model_features.pkl') # Must contain your 17 features
+
+def analyze_and_predict(headline_text, content_text, num_images, channel_tech, publish_day, num_videos, num_keywords, num_hrefs          ):
+    
+    # --- 1. Calculate Features from User Text ---
+    title_blob = TextBlob(headline_text)
+    content_blob = TextBlob(content_text)
+    
+    # Text-derived features:
+    n_tokens_title = len(headline_text.split())
+    n_tokens_content = len(content_text.split())
+    global_sentiment_polarity = content_blob.sentiment.polarity
+    global_subjectivity = content_blob.sentiment.subjectivity
+    title_sentiment_polarity = title_blob.sentiment.polarity
+    
+    # --- 2. Create Binary Weekday Flags ---
+    # Weekday names must match the names in your FEATURE_COLUMNS exactly
+    weekday_flags = {
+        'weekday_is_monday': 0, 'weekday_is_tuesday': 0, 'weekday_is_wednesday': 0, 
+        'weekday_is_thursday': 0, 'weekday_is_friday': 0, 'weekday_is_saturday': 0, 
+        'weekday_is_sunday': 0
+    }
+    # Set the selected day's flag to 1
+    if publish_day:
+        weekday_key = f'weekday_is_{publish_day.lower()}'
+        if weekday_key in weekday_flags:
+            weekday_flags[weekday_key] = 1
+
+    # --- 3. Compile all features into the input dictionary ---
+    input_data = {
+        # Text-derived features (5)
+        'n_tokens_title': n_tokens_title,
+        'n_tokens_content': n_tokens_content,
+        'global_sentiment_polarity': global_sentiment_polarity,
+        'global_subjectivity': global_subjectivity,
+        'title_sentiment_polarity': title_sentiment_polarity,
+
+        # Constant/User-Selected features (12)
+        'num_imgs': num_images,
+        'data_channel_is_tech': 1 if channel_tech else 0,
+        'num_videos': num_videos,
+        'num_keywords': num_keywords,
+        'num_hrefs': num_hrefs,
+        **weekday_flags # Adds all the dynamically set weekday flags
+    }
+    
+    # --- 4. Prepare for Prediction ---
+    # Ensure features are in the exact order the model expects (using FEATURE_COLUMNS)
+    X_pred = np.array([input_data[col] for col in FEATURE_COLUMNS]).reshape(1, -1)
+    
+    # 5. Predict and invert the log-transform
+    log_pred_shares = MODEL.predict(X_pred)[0]
+    predicted_shares = np.expm1(log_pred_shares)
+    
+    # 6. Return the results
+    return (
+        f"~{int(predicted_shares):,}", 
+        f"{global_sentiment_polarity:.3f}", 
+        f"{global_subjectivity:.3f}", 
+        f"{n_tokens_content}",
+        f"{n_tokens_title}" # New returned value
+    )
+# (Code to define analyze_and_predict is now in the app)
+# (Assuming the model loading and analyze_and_predict function are defined above)
+import streamlit as st
+
+st.title("Headline Impact: Live Popularity Predictor")
+st.markdown("Use this tool to test how your article's features affect its predicted share count.")
+
+# ----------------------------------------------------
+# --- SIDEBAR FOR CONSTANT FEATURES (Non-Text Inputs) ---
+# ----------------------------------------------------
+with st.sidebar:
+    st.header("Structural & Temporal Inputs")
+
+    # Temporal Feature
+    publish_day = st.selectbox(
+        "Day of Publication", 
+        ('Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'),
+        index=0
+    )
+    
+    # Channel Feature
+    channel_tech = st.checkbox("Is it a **Tech** Channel Article?", value=False)
+
+    st.subheader("Multimedia & Linking")
+    # Multimedia Features
+    num_images = st.slider("Number of Images (num_imgs)", 
+                           min_value=0, max_value=20, value=5, step=1)
+    num_videos = st.slider("Number of Videos (num_videos)", 
+                           min_value=0, max_value=10, value=1, step=1)
+
+    st.subheader("Article Structure")
+    # Linking/Keyword Features
+    num_hrefs = st.slider("Number of Links (num_hrefs)", 
+                          min_value=0, max_value=30, value=5, step=1)
+    num_keywords = st.slider("Number of Keywords (num_keywords)", 
+                             min_value=1, max_value=10, value=5, step=1)
+
+# ----------------------------------------------------
+# --- MAIN AREA FOR TEXT INPUTS (Defining the missing variables) ---
+# ----------------------------------------------------
+
+st.header("Article Content")
+
+# *** HEADLINE TEXT DEFINITION ***
+headline_text = st.text_input("Headline Text", 
+                              placeholder="E.g., Revolutionary AI Tool Boosts Productivity")
+
+# *** CONTENT TEXT DEFINITION ***
+content_text = st.text_area("Article Snippet (for Sentiment Analysis)", 
+                            placeholder="Paste a few paragraphs of the article content here.")
+
+# ----------------------------------------------------
+# --- PREDICTION LOGIC ---
+# ----------------------------------------------------
+
+if st.button("Analyze & Predict Shares"):
+    if headline_text and content_text:
+        
+        # 1. Input Processing
+        # The weekday_is_... flag is handled inside the analyze_and_predict function
+        
+        # 2. Call the updated function with all arguments
+        predicted_shares, polarity, subjectivity, content_length, title_length = analyze_and_predict(
+            headline_text, 
+            content_text, 
+            num_images, 
+            channel_tech, 
+            publish_day, 
+            num_videos, 
+            num_keywords, 
+            num_hrefs
+        )
+        
+        # 3. Display Results
+        st.success(f"### Predicted Shares: {predicted_shares}")
+
+        col1, col2, col3, col4 = st.columns(4)
+        col1.metric("Content Polarity", polarity)
+        col2.metric("Content Subjectivity", subjectivity)
+        col3.metric("Content Word Count", content_length)
+        col4.metric("Title Word Count", title_length)
+    else:
+        st.warning("Please enter both a Headline and an Article Snippet to run the analysis.")

popularity_model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1e47044bad89f8072048d5173868e8423c68f15f62ea842c11484874ba0158e
+size 1400

requirements.txt

@@ -1,3 +1,5 @@
-altair
-pandas
-streamlit
+streamlit
+pandas
+scikit-learn
+numpy
+textblob