Update app.py

app.py

@@ -10,6 +10,7 @@ import time
 import requests
 from collections import Counter
 import numpy as np
+from streamlit.runtime.scriptrunner import add_script_run_ctx
 
 st.set_page_config(page_title="HF Contributions", layout="wide", initial_sidebar_state="expanded")
 
@@ -213,9 +214,9 @@ def get_user_commit_stats(username):
             "score": 0
         }
 
-# Enhanced function to get trending accounts with commit-based ranking
+# Enhanced function to get trending accounts (simplified without commit fetching)
 @st.cache_data(ttl=3600)  # Cache for 1 hour
-def get_trending_accounts_with_commits(limit=100):
+def get_trending_accounts_simple(limit=100):
     try:
         # First, get top accounts by model/space count
         spaces_response = requests.get("https://huggingface.co/api/spaces", 
@@ -264,30 +265,23 @@ def get_trending_accounts_with_commits(limit=100):
         for owner, _ in top_model_owners[:100]:
             unique_users.add(owner)
         
-        # Create progress bar for fetching commit stats
-        progress_text = st.empty()
-        progress_bar = st.progress(0)
-        progress_text.text(f"Analyzing top contributors... (0/{len(unique_users)})")
-        
-        # Fetch commit stats for all unique users
+        # For now, use a simple ranking based on total repos
+        # (We'll fetch commits only for the selected user)
         user_stats = []
-        with ThreadPoolExecutor(max_workers=5) as executor:
-            future_to_user = {executor.submit(get_user_commit_stats, user): user for user in unique_users}
-            completed = 0
-            for future in as_completed(future_to_user):
-                stats = future.result()
-                if stats["score"] > 0:  # Only include users with some activity
-                    user_stats.append(stats)
-                completed += 1
-                progress = completed / len(unique_users)
-                progress_bar.progress(progress)
-                progress_text.text(f"Analyzing top contributors... ({completed}/{len(unique_users)})")
-        
-        # Clear progress indicators
-        progress_text.empty()
-        progress_bar.empty()
+        for user in unique_users:
+            models_count = next((count for owner, count in top_model_owners if owner == user), 0)
+            spaces_count = next((count for owner, count in top_space_owners if owner == user), 0)
+            
+            user_stats.append({
+                "username": user,
+                "models": models_count,
+                "spaces": spaces_count,
+                "datasets": 0,  # We'll skip datasets for initial ranking
+                "estimated_commits": 0,  # Will be calculated later for selected user
+                "score": models_count + spaces_count  # Temporary score for initial display
+            })
         
-        # Sort by score (commits only) for overall ranking
+        # Sort by total repo count for initial display
         user_stats.sort(key=lambda x: x["score"], reverse=True)
         
         # Extract rankings
@@ -358,10 +352,12 @@ def get_commit_events(username, kind=None, selected_year=None):
             for i in range(0, len(repo_ids), chunk_size):
                 chunk = repo_ids[i:i + chunk_size]
                 with ThreadPoolExecutor(max_workers=min(5, len(chunk))) as executor:
-                    future_to_repo = {
-                        executor.submit(fetch_commits_for_repo, repo_id, k, username, selected_year): repo_id
-                        for repo_id in chunk
-                    }
+                    future_to_repo = {}
+                    for repo_id in chunk:
+                        future = executor.submit(fetch_commits_for_repo, repo_id, k, username, selected_year)
+                        add_script_run_ctx(future)
+                        future_to_repo[future] = repo_id
+                    
                     for future in as_completed(future_to_repo):
                         repo_commits, repo_count = future.result()
                         if repo_commits:  # Only extend if we got commits
@@ -692,84 +688,9 @@ def simulate_follower_data(username, spaces_count, models_count, total_commits):
     
     return fig
 
-# Function to create ranking position visualization
-def create_ranking_chart(username, overall_rank, spaces_rank, models_rank):
-    if not (overall_rank or spaces_rank or models_rank):
-        return None
-    
-    # Create a horizontal bar chart for rankings with improved styling
-    fig, ax = plt.subplots(figsize=(12, 5), facecolor='#F8F9FA')
-    
-    categories = []
-    positions = []
-    colors = []
-    rank_values = []
-    
-    if overall_rank:
-        categories.append('Overall')
-        positions.append(101 - overall_rank)  # Invert rank for visualization (higher is better)
-        colors.append('#673AB7')
-        rank_values.append(overall_rank)
-    
-    if spaces_rank:
-        categories.append('Spaces')
-        positions.append(101 - spaces_rank)
-        colors.append('#2196F3')
-        rank_values.append(spaces_rank)
-    
-    if models_rank:
-        categories.append('Models')
-        positions.append(101 - models_rank)
-        colors.append('#FF9800')
-        rank_values.append(models_rank)
-    
-    # Create horizontal bars with enhanced styling
-    bars = ax.barh(categories, positions, color=colors, alpha=0.8, height=0.6, 
-                  edgecolor='white', linewidth=1.5)
-    
-    # Add rank values as text with improved styling
-    for i, bar in enumerate(bars):
-        ax.text(bar.get_width() + 2, bar.get_y() + bar.get_height()/2, 
-                f'Rank #{rank_values[i]}', va='center', fontsize=12, 
-                fontweight='bold', color=colors[i])
-    
-    # Set chart properties with enhanced styling
-    ax.set_xlim(0, 105)
-    ax.set_title(f"Ranking Positions for {username} (Top 100)", fontsize=18, pad=20, fontweight='bold')
-    ax.set_xlabel("Percentile (higher is better)", fontsize=14, labelpad=10)
-    
-    # Add explanatory text
-    ax.text(50, -0.6, "← Lower rank (higher number) | Higher rank (lower number) →", 
-            ha='center', va='center', fontsize=10, fontweight='bold', color='#666666')
-    
-    # Add a vertical line at 90th percentile to highlight top 10 with improved styling
-    ax.axvline(x=90, color='#FF5252', linestyle='--', alpha=0.7, linewidth=2)
-    ax.text(92, len(categories)/2, 'Top 10', color='#D32F2F', fontsize=12, 
-           rotation=90, va='center', fontweight='bold')
-    
-    # Style the chart borders and background
-    ax.spines['top'].set_visible(False)
-    ax.spines['right'].set_visible(False)
-    ax.spines['left'].set_linewidth(0.5)
-    ax.spines['bottom'].set_linewidth(0.5)
-    
-    # Adjust tick parameters for better look
-    ax.tick_params(axis='x', labelsize=12)
-    ax.tick_params(axis='y', labelsize=14, pad=5)
-    
-    # Add grid for better readability
-    ax.grid(axis='x', linestyle='--', alpha=0.5, color='#CCCCCC')
-    ax.set_axisbelow(True)  # Grid lines behind bars
-    
-    # Invert x-axis to show ranking position more intuitively
-    ax.invert_xaxis()
-    
-    plt.tight_layout()
-    return fig
-
 # Fetch trending accounts with a loading spinner (do this once at the beginning)
-with st.spinner("Loading and analyzing top contributors... This may take a few moments."):
-    trending_accounts, top_owners_spaces, top_owners_models, user_stats = get_trending_accounts_with_commits(limit=100)
+with st.spinner("Loading top contributors..."):
+    trending_accounts, top_owners_spaces, top_owners_models, user_stats = get_trending_accounts_simple(limit=100)
 
 # Sidebar
 with st.sidebar:
@@ -777,14 +698,14 @@ with st.sidebar:
     
     # Create tabs for rankings
     tab1, tab2 = st.tabs([
-        "Top 100 Overall", 
-        "Top Spaces & Models"
+        "Top Contributors", 
+        "Repository Rankings"
     ])
     
     with tab1:
-        # Show combined trending accounts list with commit-based ranking
-        st.markdown('<div class="subheader"><h3>🔥 Top 100 Contributors by Commits</h3></div>', unsafe_allow_html=True)
-        st.markdown('<p style="font-size: 0.9rem; color: #666; margin-bottom: 10px;">Ranked by total commit count</p>', unsafe_allow_html=True)
+        # Show combined trending accounts list
+        st.markdown('<div class="subheader"><h3>🔥 Top 100 Contributors</h3></div>', unsafe_allow_html=True)
+        st.markdown('<p style="font-size: 0.9rem; color: #666; margin-bottom: 10px;">Initial ranking by total repositories. Select a user to see commit-based analysis.</p>', unsafe_allow_html=True)
         
         # Create a data frame for the table
         if user_stats:
@@ -802,15 +723,14 @@ with st.sidebar:
                 
                 overall_data.append([
                     f"{rank_display}{stat['username']}", 
-                    str(stat['estimated_commits']),
                     str(stat['models']),
                     str(stat['spaces']),
-                    str(stat['datasets'])
+                    str(stat['models'] + stat['spaces'])  # Total repos
                 ])
             
             ranking_data_overall = pd.DataFrame(
                 overall_data, 
-                columns=["Contributor", "Total Commits", "Models", "Spaces", "Datasets"]
+                columns=["Contributor", "Models", "Spaces", "Total Repos"]
             )
             ranking_data_overall.index = ranking_data_overall.index + 1  # Start index from 1 for ranking
 
@@ -819,10 +739,9 @@ with st.sidebar:
                 height=900,  # 약 30행 정도 보이도록 픽셀 단위 높이 설정
                 column_config={
                     "Contributor": st.column_config.TextColumn("Contributor"),
-                    "Total Commits": st.column_config.TextColumn("Total Commits"),
                     "Models": st.column_config.TextColumn("Models"),
                     "Spaces": st.column_config.TextColumn("Spaces"),
-                    "Datasets": st.column_config.TextColumn("Datasets")
+                    "Total Repos": st.column_config.TextColumn("Total Repos")
                 },
                 use_container_width=True,
                 hide_index=False
@@ -939,11 +858,9 @@ if username:
     # Create a header card with contributor info
     header_col1, header_col2 = st.columns([1, 2])
     with header_col1:
-        commits_display = f"Est. Commits: {user_stat['estimated_commits']}" if user_stat else "Est. Commits: N/A"
         st.markdown(f'<div style="background-color: #E3F2FD; padding: 20px; border-radius: 10px; border-left: 5px solid #1E88E5;">'
                   f'<h2 style="color: #1E88E5;">👤 {username}</h2>'
                   f'<p style="font-size: 16px;">Analyzing contributions for {selected_year}</p>'
-                  f'<p style="font-size: 14px; font-weight: bold;">{commits_display}</p>'
                   f'<p><a href="https://huggingface.co/{username}" target="_blank" style="color: #1E88E5; font-weight: bold;">View Profile</a></p>'
                   f'</div>', unsafe_allow_html=True)
     
@@ -952,7 +869,7 @@ if username:
         st.markdown(f'<div style="background-color: #F3E5F5; padding: 20px; border-radius: 10px; border-left: 5px solid #9C27B0;">'
                   f'<h3 style="color: #9C27B0;">About This Analysis</h3>'
                   f'<p>This dashboard analyzes {username}\'s contributions to Hugging Face in {selected_year}, including models, datasets, and spaces.</p>'
-                  f'<p style="font-style: italic; font-size: 12px;">* Overall rankings are based on total commit count. Space/Model rankings are based on repository count.</p>'
+                  f'<p style="font-style: italic; font-size: 12px;">* Detailed commit analysis will be calculated after selection.</p>'
                   f'</div>', unsafe_allow_html=True)
 
     with st.spinner(f"Fetching detailed contribution data for {username}..."):
@@ -970,7 +887,8 @@ if username:
             
             # Create a prominent ranking display
             st.markdown(f'<div style="background-color: #FFF8E1; padding: 20px; border-radius: 10px; border-left: 5px solid #FFC107; margin: 1rem 0;">'
-                      f'<h2 style="color: #FFA000; text-align: center;">🏆 Ranked #{overall_rank} in Top Contributors</h2>'
+                      f'<h2 style="color: #FFA000; text-align: center;">📊 Featured Contributor</h2>'
+                      f'<p style="text-align: center; font-size: 16px;">Analyzing detailed commit statistics...</p>'
                       f'</div>', unsafe_allow_html=True)
             
             # Find user in spaces ranking
@@ -987,10 +905,56 @@ if username:
                     models_count = count
                     break
                     
-            # Display ranking visualization
-            rank_chart = create_ranking_chart(username, overall_rank, spaces_rank, models_rank)
-            if rank_chart:
-                st.pyplot(rank_chart)
+            # Display ranking visualization only if user has rankings
+            if spaces_rank or models_rank:
+                # Create custom ranking chart for spaces and models only
+                fig, ax = plt.subplots(figsize=(12, 5), facecolor='#F8F9FA')
+                
+                categories = []
+                positions = []
+                colors = []
+                rank_values = []
+                
+                if spaces_rank:
+                    categories.append('Spaces')
+                    positions.append(101 - spaces_rank)
+                    colors.append('#2196F3')
+                    rank_values.append(spaces_rank)
+                
+                if models_rank:
+                    categories.append('Models')
+                    positions.append(101 - models_rank)
+                    colors.append('#FF9800')
+                    rank_values.append(models_rank)
+                
+                if categories:  # Only create chart if there are rankings
+                    bars = ax.barh(categories, positions, color=colors, alpha=0.8, height=0.6, 
+                                  edgecolor='white', linewidth=1.5)
+                    
+                    for i, bar in enumerate(bars):
+                        ax.text(bar.get_width() + 2, bar.get_y() + bar.get_height()/2, 
+                                f'Rank #{rank_values[i]}', va='center', fontsize=12, 
+                                fontweight='bold', color=colors[i])
+                    
+                    ax.set_xlim(0, 105)
+                    ax.set_title(f"Repository Rankings for {username} (Top 100)", fontsize=18, pad=20, fontweight='bold')
+                    ax.set_xlabel("Percentile (higher is better)", fontsize=14, labelpad=10)
+                    
+                    ax.spines['top'].set_visible(False)
+                    ax.spines['right'].set_visible(False)
+                    ax.spines['left'].set_linewidth(0.5)
+                    ax.spines['bottom'].set_linewidth(0.5)
+                    
+                    ax.tick_params(axis='x', labelsize=12)
+                    ax.tick_params(axis='y', labelsize=14, pad=5)
+                    
+                    ax.grid(axis='x', linestyle='--', alpha=0.5, color='#CCCCCC')
+                    ax.set_axisbelow(True)
+                    
+                    ax.invert_xaxis()
+                    
+                    plt.tight_layout()
+                    st.pyplot(fig)
         
         # Create a dictionary to store commits by type
         commits_by_type = {}
@@ -1163,12 +1127,16 @@ if username:
                           f'</ul>', unsafe_allow_html=True)
                 
                 # Add ranking context if available
-                if overall_rank:
-                    percentile = 100 - overall_rank
+                if spaces_rank or models_rank:
                     st.markdown(f'<div style="margin-top: 20px;">'
-                              f'<h3 style="color: #1E88E5; border-bottom: 1px solid #E0E0E0; padding-bottom: 10px;">Ranking Analysis</h3>'
-                              f'<ul style="list-style-type: none; padding-left: 5px;">'
-                              f'<li style="margin: 15px 0; font-size: 16px;">🏆 <strong>Overall Ranking:</strong> #{overall_rank} (Top {percentile}% of contributors)</li>', unsafe_allow_html=True)
+                              f'<h3 style="color: #1E88E5; border-bottom: 1px solid #E0E0E0; padding-bottom: 10px;">Repository Rankings</h3>'
+                              f'<ul style="list-style-type: none; padding-left: 5px;">', unsafe_allow_html=True)
+                    
+                    if spaces_rank:
+                        st.markdown(f'<li style="margin: 15px 0; font-size: 16px;">🚀 <strong>Spaces Ranking:</strong> #{spaces_rank} with {spaces_count} spaces</li>', unsafe_allow_html=True)
+                    
+                    if models_rank:
+                        st.markdown(f'<li style="margin: 15px 0; font-size: 16px;">🧠 <strong>Models Ranking:</strong> #{models_rank} with {models_count} models</li>', unsafe_allow_html=True)
                     
                     badge_html = '<div style="margin: 20px 0;">'