gitflow-analytics 1.0.3__py3-none-any.whl → 1.3.6__py3-none-any.whl

gitflow_analytics/reports/analytics_writer.py

@@ -1,26 +1,232 @@
 """Advanced analytics report generation with percentage and qualitative metrics."""
 import csv
+import logging
+from collections import defaultdict
 from datetime import datetime, timedelta, timezone
 from pathlib import Path
-from typing import List, Dict, Any, Tuple
-from collections import defaultdict
-import pandas as pd
+from typing import Any, Dict, List, Tuple
+
 import numpy as np
+import pandas as pd
+
+# Get logger for this module
+logger = logging.getLogger(__name__)
 
 
 class AnalyticsReportGenerator:
     """Generate advanced analytics reports with percentage breakdowns and qualitative insights."""
     
-    def __init__(self, anonymize: bool = False):
+    def __init__(self, anonymize: bool = False, exclude_authors: list[str] = None, identity_resolver=None):
         """Initialize analytics report generator."""
         self.anonymize = anonymize
         self._anonymization_map = {}
         self._anonymous_counter = 0
+        self.exclude_authors = exclude_authors or []
+        self.identity_resolver = identity_resolver
+
+    def _filter_excluded_authors(self, data_list: list[dict[str, Any]]) -> list[dict[str, Any]]:
+        """
+        Filter out excluded authors from any data list using canonical_id.
+        
+        WHY: Bot exclusion happens in Phase 2 (reporting) instead of Phase 1 (data collection)
+        to ensure manual identity mappings work correctly. This allows the system to see 
+        consolidated bot identities via canonical_id instead of just original author_email/author_name.
+        
+        Args:
+            data_list: List of data dictionaries containing canonical_id field
+            
+        Returns:
+            Filtered list with excluded authors removed
+        """
+        if not self.exclude_authors:
+            return data_list
+            
+        logger.debug(f"DEBUG EXCLUSION: Starting filter with {len(self.exclude_authors)} excluded authors: {self.exclude_authors}")
+        logger.debug(f"DEBUG EXCLUSION: Filtering {len(data_list)} items from data list")
+        
+        excluded_lower = [author.lower() for author in self.exclude_authors]
+        logger.debug(f"DEBUG EXCLUSION: Excluded authors (lowercase): {excluded_lower}")
+        
+        filtered_data = []
+        excluded_count = 0
+        
+        # Sample first 5 items to see data structure
+        for i, item in enumerate(data_list[:5]):
+            logger.debug(f"DEBUG EXCLUSION: Sample item {i}: canonical_id='{item.get('canonical_id', '')}', "
+                        f"author_email='{item.get('author_email', '')}', author_name='{item.get('author_name', '')}', "
+                        f"author='{item.get('author', '')}', primary_name='{item.get('primary_name', '')}', "
+                        f"name='{item.get('name', '')}', developer='{item.get('developer', '')}', "
+                        f"display_name='{item.get('display_name', '')}'")  
+        
+        for item in data_list:
+            canonical_id = item.get("canonical_id", "")
+            # Also check original author fields as fallback for data without canonical_id
+            author_email = item.get("author_email", "")
+            author_name = item.get("author_name", "")
+            
+            # Check all possible author fields to ensure we catch every variation
+            author = item.get("author", "")
+            primary_name = item.get("primary_name", "")
+            name = item.get("name", "")
+            developer = item.get("developer", "")  # Common in analytics data
+            display_name = item.get("display_name", "")  # Common in some data structures
+            
+            # Check canonical_id FIRST - this is the primary exclusion check
+            should_exclude = False
+            if canonical_id and canonical_id.lower() in excluded_lower:
+                should_exclude = True
+            # CRITICAL: Also check primary_email for manual mappings (e.g. bots mapped to bot@excluded.local)
+            elif item.get("primary_email", "") and item.get("primary_email", "").lower() in excluded_lower:
+                should_exclude = True
+            # Fall back to checking other fields only if canonical_id and primary_email don't match
+            elif not should_exclude:
+                should_exclude = (
+                    (author_email and author_email.lower() in excluded_lower) or
+                    (author_name and author_name.lower() in excluded_lower) or
+                    (author and author.lower() in excluded_lower) or
+                    (primary_name and primary_name.lower() in excluded_lower) or
+                    (name and name.lower() in excluded_lower) or
+                    (developer and developer.lower() in excluded_lower) or
+                    (display_name and display_name.lower() in excluded_lower)
+                )
+            
+            if should_exclude:
+                excluded_count += 1
+                logger.debug(f"DEBUG EXCLUSION: EXCLUDING item - canonical_id='{canonical_id}', "
+                           f"primary_email='{item.get('primary_email', '')}', "
+                           f"author_email='{author_email}', author_name='{author_name}', author='{author}', "
+                           f"primary_name='{primary_name}', name='{name}', developer='{developer}', "
+                           f"display_name='{display_name}'")
+            else:
+                filtered_data.append(item)
+                
+        logger.debug(f"DEBUG EXCLUSION: Excluded {excluded_count} items, kept {len(filtered_data)} items")
+        return filtered_data
+
+    def _get_canonical_display_name(self, canonical_id: str, fallback_name: str) -> str:
+        """
+        Get the canonical display name for a developer.
+        
+        WHY: Manual identity mappings may have updated display names that aren't
+        reflected in the developer_stats data passed to report generators. This
+        method ensures we get the most current display name from the identity resolver.
+        
+        Args:
+            canonical_id: The canonical ID to get the display name for
+            fallback_name: The fallback name to use if identity resolver is not available
+            
+        Returns:
+            The canonical display name or fallback name
+        """
+        if self.identity_resolver and canonical_id:
+            try:
+                canonical_name = self.identity_resolver.get_canonical_name(canonical_id)
+                if canonical_name and canonical_name != "Unknown":
+                    return canonical_name
+            except Exception as e:
+                logger.debug(f"Error getting canonical name for {canonical_id}: {e}")
+        
+        return fallback_name
+    
+    def _get_files_changed_count(self, commit: Dict[str, Any]) -> int:
+        """Safely extract files_changed count from commit data.
+        
+        WHY: The files_changed field can be either an int (count) or list (file names).
+        This helper ensures we always get an integer count for calculations.
+        
+        Args:
+            commit: Commit dictionary with files_changed field
+            
+        Returns:
+            Integer count of files changed
+        """
+        files_changed = commit.get('files_changed', 0)
+        
+        if isinstance(files_changed, int):
+            return files_changed
+        elif isinstance(files_changed, list):
+            return len(files_changed)
+        else:
+            # Fallback for unexpected types
+            logger.warning(f"Unexpected files_changed type: {type(files_changed)}, defaulting to 0")
+            return 0
+    
+    def _log_datetime_comparison(self, dt1: datetime, dt2: datetime, operation: str, location: str) -> None:
+        """Log datetime comparison details for debugging timezone issues."""
+        logger.debug(f"Comparing dates in {location} ({operation}):")
+        logger.debug(f"  dt1: {dt1} (tzinfo: {dt1.tzinfo}, aware: {dt1.tzinfo is not None})")
+        logger.debug(f"  dt2: {dt2} (tzinfo: {dt2.tzinfo}, aware: {dt2.tzinfo is not None})")
+    
+    def _safe_datetime_compare(self, dt1: datetime, dt2: datetime, operation: str, location: str) -> bool:
+        """Safely compare datetimes with logging and error handling."""
+        try:
+            self._log_datetime_comparison(dt1, dt2, operation, location)
+            
+            if operation == 'lt':
+                result = dt1 < dt2
+            elif operation == 'gt':
+                result = dt1 > dt2
+            elif operation == 'le':
+                result = dt1 <= dt2
+            elif operation == 'ge':
+                result = dt1 >= dt2
+            elif operation == 'eq':
+                result = dt1 == dt2
+            else:
+                raise ValueError(f"Unknown operation: {operation}")
+                
+            logger.debug(f"  Result: {result}")
+            return result
+            
+        except TypeError as e:
+            logger.error(f"Timezone comparison error in {location}:")
+            logger.error(f"  dt1: {dt1} (type: {type(dt1)}, tzinfo: {getattr(dt1, 'tzinfo', 'N/A')})")
+            logger.error(f"  dt2: {dt2} (type: {type(dt2)}, tzinfo: {getattr(dt2, 'tzinfo', 'N/A')})")
+            logger.error(f"  Operation: {operation}")
+            logger.error(f"  Error: {e}")
+            
+            # Import traceback for detailed error info
+            import traceback
+            logger.error(f"  Full traceback:\n{traceback.format_exc()}")
+            
+            # Try to fix by making both timezone-aware in UTC
+            try:
+                if dt1.tzinfo is None:
+                    dt1 = dt1.replace(tzinfo=timezone.utc)
+                    logger.debug(f"  Fixed dt1 to UTC: {dt1}")
+                if dt2.tzinfo is None:
+                    dt2 = dt2.replace(tzinfo=timezone.utc)
+                    logger.debug(f"  Fixed dt2 to UTC: {dt2}")
+                
+                # Retry comparison
+                if operation == 'lt':
+                    result = dt1 < dt2
+                elif operation == 'gt':
+                    result = dt1 > dt2
+                elif operation == 'le':
+                    result = dt1 <= dt2
+                elif operation == 'ge':
+                    result = dt1 >= dt2
+                elif operation == 'eq':
+                    result = dt1 == dt2
+                else:
+                    raise ValueError(f"Unknown operation: {operation}")
+                    
+                logger.info(f"  Fixed comparison result: {result}")
+                return result
+                
+            except Exception as fix_error:
+                logger.error(f"  Failed to fix timezone issue: {fix_error}")
+                raise
     
     def generate_activity_distribution_report(self, commits: List[Dict[str, Any]], 
                                             developer_stats: List[Dict[str, Any]],
                                             output_path: Path) -> Path:
         """Generate activity distribution report with percentage breakdowns."""
+        # Apply exclusion filtering in Phase 2
+        commits = self._filter_excluded_authors(commits)
+        developer_stats = self._filter_excluded_authors(developer_stats)
+        
         # Build lookup maps
         dev_lookup = {dev['canonical_id']: dev for dev in developer_stats}
         
@@ -31,7 +237,7 @@ class AnalyticsReportGenerator:
             c.get('filtered_deletions', c.get('deletions', 0)) 
             for c in commits
         )
-        total_files = sum(c['files_changed'] for c in commits)
+        total_files = sum(self._get_files_changed_count(c) for c in commits)
         
         # Group by developer and project
         dev_project_activity = defaultdict(lambda: defaultdict(lambda: {
@@ -47,7 +253,16 @@ class AnalyticsReportGenerator:
                 commit.get('filtered_insertions', commit.get('insertions', 0)) + 
                 commit.get('filtered_deletions', commit.get('deletions', 0))
             )
-            dev_project_activity[dev_id][project]['files'] += commit.get('filtered_files_changed', commit.get('files_changed', 0))
+            # Handle files_changed safely - could be int or list
+            files_changed = commit.get('filtered_files_changed')
+            if files_changed is None:
+                files_changed = self._get_files_changed_count(commit)
+            elif isinstance(files_changed, list):
+                files_changed = len(files_changed)
+            elif not isinstance(files_changed, int):
+                files_changed = 0
+            
+            dev_project_activity[dev_id][project]['files'] += files_changed
             dev_project_activity[dev_id][project]['story_points'] += commit.get('story_points', 0) or 0
         
         # Build report data
@@ -55,7 +270,12 @@ class AnalyticsReportGenerator:
         
         for dev_id, projects in dev_project_activity.items():
             developer = dev_lookup.get(dev_id, {})
-            dev_name = self._anonymize_value(developer.get('primary_name', 'Unknown'), 'name')
+            dev_name = self._anonymize_value(
+                self._get_canonical_display_name(
+                    dev_id, 
+                    developer.get('primary_name', 'Unknown')
+                ), 'name'
+            )
             
             # Calculate developer totals
             dev_total_commits = sum(p['commits'] for p in projects.values())
@@ -98,6 +318,9 @@ class AnalyticsReportGenerator:
                                            ticket_analysis: Dict[str, Any],
                                            output_path: Path) -> Path:
         """Generate qualitative insights and patterns report."""
+        # Apply exclusion filtering in Phase 2
+        commits = self._filter_excluded_authors(commits)
+        developer_stats = self._filter_excluded_authors(developer_stats)
         insights = []
         
         # Analyze commit patterns
@@ -127,10 +350,18 @@ class AnalyticsReportGenerator:
                                       output_path: Path,
                                       weeks: int = 12) -> Path:
         """Generate developer focus analysis showing concentration patterns and activity across all projects."""
+        # Apply exclusion filtering in Phase 2
+        commits = self._filter_excluded_authors(commits)
+        developer_stats = self._filter_excluded_authors(developer_stats)
+        
         # Calculate week boundaries (timezone-aware to match commit timestamps)
         end_date = datetime.now(timezone.utc)
         start_date = end_date - timedelta(weeks=weeks)
         
+        logger.debug(f"Developer focus report date range:")
+        logger.debug(f"  start_date: {start_date} (tzinfo: {start_date.tzinfo})")
+        logger.debug(f"  end_date: {end_date} (tzinfo: {end_date.tzinfo})")
+        
         # Build developer lookup
         dev_lookup = {dev['canonical_id']: dev for dev in developer_stats}
         
@@ -149,7 +380,12 @@ class AnalyticsReportGenerator:
         
         for dev in developer_stats:
             dev_id = dev['canonical_id']
-            dev_name = self._anonymize_value(dev['primary_name'], 'name')
+            dev_name = self._anonymize_value(
+                self._get_canonical_display_name(
+                    dev_id, 
+                    dev['primary_name']
+                ), 'name'
+            )
             
             # Get developer's commits
             dev_commits = [c for c in commits if c.get('canonical_id') == dev_id]
@@ -164,6 +400,10 @@ class AnalyticsReportGenerator:
             commit_hours = []
             
             for commit in dev_commits:
+                # Log commit processing
+                logger.debug(f"Processing commit for developer {dev_name}: {commit.get('hash', 'unknown')[:8]}")
+                logger.debug(f"  timestamp: {commit['timestamp']} (tzinfo: {getattr(commit['timestamp'], 'tzinfo', 'N/A')})")
+                
                 # Project distribution
                 project_key = commit.get('project_key', 'UNKNOWN')
                 projects[project_key] += 1
@@ -182,8 +422,10 @@ class AnalyticsReportGenerator:
                 # Commit size
                 commit_sizes.append(lines_changed)
                 
-                # Time of day
-                if hasattr(commit['timestamp'], 'hour'):
+                # Time of day (use local hour if available, fallback to UTC)
+                if 'local_hour' in commit:
+                    commit_hours.append(commit['local_hour'])
+                elif hasattr(commit['timestamp'], 'hour'):
                     commit_hours.append(commit['timestamp'].hour)
             
             # Calculate metrics
@@ -276,12 +518,284 @@ class AnalyticsReportGenerator:
         
         return output_path
     
+    def generate_weekly_trends_report(self, commits: List[Dict[str, Any]], 
+                                    developer_stats: List[Dict[str, Any]],
+                                    output_path: Path,
+                                    weeks: int = 12) -> Path:
+        """Generate weekly trends analysis showing changes in activity patterns."""
+        # Apply exclusion filtering in Phase 2
+        commits = self._filter_excluded_authors(commits)
+        developer_stats = self._filter_excluded_authors(developer_stats)
+        
+        # Calculate week boundaries
+        end_date = datetime.now(timezone.utc)
+        start_date = end_date - timedelta(weeks=weeks)
+        
+        # Build developer lookup
+        dev_lookup = {dev['canonical_id']: dev for dev in developer_stats}
+        
+        # Initialize data structures
+        weekly_data = defaultdict(lambda: {
+            'commits': 0,
+            'developers': set(),
+            'projects': defaultdict(int),
+            'lines_changed': 0,
+            'story_points': 0
+        })
+        
+        developer_weekly = defaultdict(lambda: defaultdict(lambda: {
+            'commits': 0, 'lines': 0, 'story_points': 0
+        }))
+        project_weekly = defaultdict(lambda: defaultdict(lambda: {
+            'commits': 0, 'lines': 0, 'developers': set(), 'story_points': 0
+        }))
+        
+        # Process commits
+        for commit in commits:
+            week_start = self._get_week_start(commit['timestamp'])
+            week_key = week_start.strftime('%Y-%m-%d')
+            
+            # Overall weekly metrics
+            weekly_data[week_key]['commits'] += 1
+            weekly_data[week_key]['developers'].add(commit.get('canonical_id'))
+            weekly_data[week_key]['projects'][commit.get('project_key', 'UNKNOWN')] += 1
+            lines = (
+                commit.get('filtered_insertions', commit.get('insertions', 0)) + 
+                commit.get('filtered_deletions', commit.get('deletions', 0))
+            )
+            weekly_data[week_key]['lines_changed'] += lines
+            weekly_data[week_key]['story_points'] += commit.get('story_points', 0) or 0
+            
+            # Developer-specific weekly data
+            dev_id = commit.get('canonical_id')
+            developer_weekly[dev_id][week_key]['commits'] += 1
+            developer_weekly[dev_id][week_key]['lines'] += lines
+            developer_weekly[dev_id][week_key]['story_points'] += commit.get('story_points', 0) or 0
+            
+            # Project-specific weekly data
+            project = commit.get('project_key', 'UNKNOWN')
+            project_weekly[project][week_key]['commits'] += 1
+            project_weekly[project][week_key]['lines'] += lines
+            project_weekly[project][week_key]['developers'].add(dev_id)
+            project_weekly[project][week_key]['story_points'] += commit.get('story_points', 0) or 0
+        
+        # Convert to rows for CSV
+        rows = []
+        sorted_weeks = sorted(weekly_data.keys())
+        
+        # Track developer and project trends
+        dev_activity_changes = defaultdict(list)  # dev_id -> list of weekly changes
+        project_activity_changes = defaultdict(list)  # project -> list of weekly changes
+        
+        for i, week in enumerate(sorted_weeks):
+            data = weekly_data[week]
+            
+            # Calculate week-over-week changes
+            prev_week = sorted_weeks[i-1] if i > 0 else None
+            
+            commits_change = 0
+            developers_change = 0
+            if prev_week:
+                prev_data = weekly_data[prev_week]
+                commits_change = data['commits'] - prev_data['commits']
+                developers_change = len(data['developers']) - len(prev_data['developers'])
+            
+            # Top project and developer this week
+            top_project = max(data['projects'].items(), key=lambda x: x[1])[0] if data['projects'] else 'NONE'
+            
+            # Find top developer this week
+            top_dev_id = None
+            top_dev_commits = 0
+            for dev_id in data['developers']:
+                dev_commits = developer_weekly[dev_id][week]['commits']
+                if dev_commits > top_dev_commits:
+                    top_dev_commits = dev_commits
+                    top_dev_id = dev_id
+            
+            top_dev_name = self._anonymize_value(
+                self._get_canonical_display_name(
+                    top_dev_id, 
+                    dev_lookup.get(top_dev_id, {}).get('primary_name', 'Unknown')
+                ), 'name'
+            ) if top_dev_id else 'None'
+            
+            # Calculate developer trends for active developers this week
+            dev_trend_summary = []
+            for dev_id in data['developers']:
+                dev_data = developer_weekly[dev_id][week]
+                prev_dev_data = developer_weekly[dev_id].get(prev_week, {'commits': 0}) if prev_week else {'commits': 0}
+                change = dev_data['commits'] - prev_dev_data['commits']
+                if change != 0:
+                    dev_name = self._anonymize_value(
+                        self._get_canonical_display_name(
+                            dev_id, 
+                            dev_lookup.get(dev_id, {}).get('primary_name', 'Unknown')
+                        ), 'name'
+                    )
+                    dev_activity_changes[dev_name].append(change)
+                    if abs(change) >= 3:  # Significant changes only
+                        dev_trend_summary.append(f"{dev_name}({'+' if change > 0 else ''}{change})")
+            
+            # Calculate project trends
+            project_trend_summary = []
+            for project, count in data['projects'].items():
+                prev_count = weekly_data[prev_week]['projects'].get(project, 0) if prev_week else 0
+                change = count - prev_count
+                if change != 0:
+                    project_activity_changes[project].append(change)
+                    if abs(change) >= 3:  # Significant changes only
+                        project_trend_summary.append(f"{project}({'+' if change > 0 else ''}{change})")
+            
+            row = {
+                'week_start': week,
+                'commits': data['commits'],
+                'active_developers': len(data['developers']),
+                'active_projects': len(data['projects']),
+                'lines_changed': data['lines_changed'],
+                'story_points': data['story_points'],
+                'commits_change': commits_change,
+                'developers_change': developers_change,
+                'top_project': top_project,
+                'top_developer': top_dev_name,
+                'avg_commits_per_dev': round(data['commits'] / max(len(data['developers']), 1), 1),
+                'avg_lines_per_commit': round(data['lines_changed'] / max(data['commits'], 1), 1),
+                'developer_trends': '; '.join(dev_trend_summary[:5]) if dev_trend_summary else 'stable',
+                'project_trends': '; '.join(project_trend_summary[:5]) if project_trend_summary else 'stable'
+            }
+            rows.append(row)
+        
+        # Write main CSV
+        df = pd.DataFrame(rows)
+        df.to_csv(output_path, index=False)
+        
+        # Also generate detailed developer trends CSV with weekly columns
+        dev_trends_path = output_path.parent / f'developer_trends_{output_path.stem.split("_")[-1]}.csv'
+        dev_trend_rows = []
+        
+        # Build developer activity by week
+        for dev_id, weekly_commits in developer_weekly.items():
+            dev_info = dev_lookup.get(dev_id, {})
+            dev_name = self._anonymize_value(
+                self._get_canonical_display_name(
+                    dev_id, 
+                    dev_info.get('primary_name', 'Unknown')
+                ), 'name'
+            )
+            
+            # Calculate summary statistics
+            weekly_values = []
+            for week in sorted_weeks:
+                commits = weekly_commits.get(week, {}).get('commits', 0)
+                weekly_values.append(commits)
+            
+            # Only include developers with any activity
+            if sum(weekly_values) > 0:
+                # Calculate trend metrics
+                changes = []
+                for i in range(1, len(weekly_values)):
+                    changes.append(weekly_values[i] - weekly_values[i-1])
+                
+                avg_change = sum(changes) / len(changes) if changes else 0
+                volatility = np.std(changes) if len(changes) > 1 else 0
+                trend = 'increasing' if avg_change > 1 else 'decreasing' if avg_change < -1 else 'stable'
+                
+                row = {
+                    'developer': dev_name,
+                    'total_commits': sum(weekly_values),
+                    'avg_weekly_commits': round(sum(weekly_values) / len(weekly_values), 1),
+                    'avg_weekly_change': round(avg_change, 1),
+                    'volatility': round(volatility, 1),
+                    'trend': trend,
+                    'total_weeks_active': len([v for v in weekly_values if v > 0]),
+                    'max_week': max(weekly_values),
+                    'min_week': min([v for v in weekly_values if v > 0]) if any(v > 0 for v in weekly_values) else 0
+                }
+                
+                # Add weekly columns
+                for i, week in enumerate(sorted_weeks):
+                    week_label = f'week_{i+1}_{week}'
+                    row[week_label] = weekly_values[i]
+                
+                dev_trend_rows.append(row)
+        
+        if dev_trend_rows:
+            dev_trends_df = pd.DataFrame(dev_trend_rows)
+            # Sort by total commits to show most active developers first
+            dev_trends_df.sort_values('total_commits', ascending=False, inplace=True)
+            dev_trends_df.to_csv(dev_trends_path, index=False)
+        
+        # Also generate detailed project trends CSV with weekly columns
+        proj_trends_path = output_path.parent / f'project_trends_{output_path.stem.split("_")[-1]}.csv'
+        proj_trend_rows = []
+        
+        # Build project activity by week
+        for project, weekly_commits in project_weekly.items():
+            # Calculate summary statistics
+            weekly_values = []
+            weekly_developers = []
+            for week in sorted_weeks:
+                commits = weekly_commits.get(week, {}).get('commits', 0)
+                weekly_values.append(commits)
+                # Count unique developers for this project this week
+                devs = weekly_commits.get(week, {}).get('developers', set())
+                weekly_developers.append(len(devs))
+            
+            # Only include projects with any activity
+            if sum(weekly_values) > 0:
+                # Calculate trend metrics
+                changes = []
+                for i in range(1, len(weekly_values)):
+                    changes.append(weekly_values[i] - weekly_values[i-1])
+                
+                avg_change = sum(changes) / len(changes) if changes else 0
+                volatility = np.std(changes) if len(changes) > 1 else 0
+                trend = 'growing' if avg_change > 2 else 'shrinking' if avg_change < -2 else 'stable'
+                
+                row = {
+                    'project': project,
+                    'total_commits': sum(weekly_values),
+                    'avg_weekly_commits': round(sum(weekly_values) / len(weekly_values), 1),
+                    'avg_weekly_developers': round(sum(weekly_developers) / len(weekly_developers), 1),
+                    'avg_weekly_change': round(avg_change, 1),
+                    'volatility': round(volatility, 1),
+                    'trend': trend,
+                    'total_weeks_active': len([v for v in weekly_values if v > 0]),
+                    'max_week': max(weekly_values),
+                    'min_week': min([v for v in weekly_values if v > 0]) if any(v > 0 for v in weekly_values) else 0
+                }
+                
+                # Add weekly columns for commits
+                for i, week in enumerate(sorted_weeks):
+                    week_label = f'week_{i+1}_{week}'
+                    row[week_label] = weekly_values[i]
+                
+                # Add weekly columns for developer count
+                for i, week in enumerate(sorted_weeks):
+                    week_label = f'devs_week_{i+1}'
+                    row[week_label] = weekly_developers[i]
+                
+                proj_trend_rows.append(row)
+        
+        if proj_trend_rows:
+            proj_trends_df = pd.DataFrame(proj_trend_rows)
+            # Sort by total commits to show most active projects first
+            proj_trends_df.sort_values('total_commits', ascending=False, inplace=True)
+            proj_trends_df.to_csv(proj_trends_path, index=False)
+        
+        return output_path
+    
     def _analyze_commit_patterns(self, commits: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
         """Analyze patterns in commit data."""
         insights = []
         
-        # Time-based patterns
-        commit_hours = [c['timestamp'].hour for c in commits if hasattr(c['timestamp'], 'hour')]
+        # Time-based patterns (use local hour if available)
+        commit_hours = []
+        for c in commits:
+            if 'local_hour' in c:
+                commit_hours.append(c['local_hour'])
+            elif hasattr(c['timestamp'], 'hour'):
+                commit_hours.append(c['timestamp'].hour)
+        
         if commit_hours:
             peak_hour = max(set(commit_hours), key=commit_hours.count)
             insights.append({
@@ -410,7 +924,11 @@ class AnalyticsReportGenerator:
         insights = []
         
         # File change patterns
-        file_changes = [c['files_changed'] for c in commits if c['files_changed'] > 0]
+        file_changes = []
+        for c in commits:
+            files_count = self._get_files_changed_count(c)
+            if files_count > 0:
+                file_changes.append(files_count)
         if file_changes:
             avg_files = np.mean(file_changes)
             
@@ -442,18 +960,25 @@ class AnalyticsReportGenerator:
     
     def _get_week_start(self, date: datetime) -> datetime:
         """Get Monday of the week for a given date."""
+        logger.debug(f"Getting week start for date: {date} (tzinfo: {getattr(date, 'tzinfo', 'N/A')})")
+        
         # Ensure consistent timezone handling - keep timezone info
         if hasattr(date, 'tzinfo') and date.tzinfo is not None:
             # Keep timezone-aware but ensure it's UTC
             if date.tzinfo != timezone.utc:
                 date = date.astimezone(timezone.utc)
+                logger.debug(f"  Converted to UTC: {date}")
         else:
             # Convert naive datetime to UTC timezone-aware
             date = date.replace(tzinfo=timezone.utc)
+            logger.debug(f"  Made timezone-aware: {date}")
         
         days_since_monday = date.weekday()
         monday = date - timedelta(days=days_since_monday)
-        return monday.replace(hour=0, minute=0, second=0, microsecond=0)
+        result = monday.replace(hour=0, minute=0, second=0, microsecond=0)
+        
+        logger.debug(f"  Week start result: {result} (tzinfo: {result.tzinfo})")
+        return result
     
     def _anonymize_value(self, value: str, field_type: str) -> str:
         """Anonymize a value if anonymization is enabled."""