modius_performance/visualization/base_chart.py

"""
Base chart class for creating visualizations.
"""
import plotly.graph_objects as go
import plotly.express as px
import pandas as pd
import logging
from datetime import datetime
from typing import List, Dict, Any, Optional, Tuple
from abc import ABC, abstractmethod

from ..config.constants import CHART_CONFIG, CHART_COLORS, Y_AXIS_RANGES, FILE_PATHS
from ..data.data_processor import DataProcessor

logger = logging.getLogger(__name__)


class BaseChart(ABC):
    """Base class for all chart visualizations."""
    
    def __init__(self, data_processor: DataProcessor = None):
        self.data_processor = data_processor or DataProcessor()
        self.config = CHART_CONFIG
        self.colors = CHART_COLORS
        self.y_ranges = Y_AXIS_RANGES
        self.file_paths = FILE_PATHS
    
    @abstractmethod
    def create_chart(self, df: pd.DataFrame, **kwargs) -> go.Figure:
        """Create the chart visualization."""
        pass
    
    def _create_base_figure(self) -> go.Figure:
        """Create a base figure with common settings."""
        return go.Figure()
    
    def _add_background_shapes(self, fig: go.Figure, min_time: datetime, max_time: datetime, 
                              y_min: float, y_max: float) -> None:
        """Add background shapes for positive and negative regions."""
        # Add shape for positive region (above zero)
        fig.add_shape(
            type="rect",
            fillcolor=self.colors['positive_region'],
            line=dict(width=0),
            y0=0, y1=y_max,
            x0=min_time, x1=max_time,
            layer="below"
        )
        
        # Add shape for negative region (below zero)
        fig.add_shape(
            type="rect",
            fillcolor=self.colors['negative_region'],
            line=dict(width=0),
            y0=y_min, y1=0,
            x0=min_time, x1=max_time,
            layer="below"
        )
    
    def _add_zero_line(self, fig: go.Figure, min_time: datetime, max_time: datetime) -> None:
        """Add a zero line to the chart."""
        fig.add_shape(
            type="line",
            line=dict(dash="solid", width=1.5, color=self.colors['zero_line']),
            y0=0, y1=0,
            x0=min_time, x1=max_time
        )
    
    def _update_layout(self, fig: go.Figure, title: str, y_axis_title: str = None, 
                      height: int = None, y_range: List[float] = None) -> None:
        """Update the figure layout with common settings."""
        fig.update_layout(
            title=dict(
                text=title,
                font=dict(
                    family=self.config['font_family'],
                    size=self.config['title_size'],
                    color="black",
                    weight="bold"
                )
            ),
            xaxis_title=None,
            yaxis_title=None,
            template=self.config['template'],
            height=height or self.config['height'],
            autosize=True,
            legend=dict(
                orientation="h",
                yanchor="bottom",
                y=1.05,
                xanchor="center",
                x=0.5,
                groupclick="toggleitem",
                font=dict(
                    family=self.config['font_family'],
                    size=self.config['legend_font_size'],
                    color="black",
                    weight="bold"
                )
            ),
            margin=dict(r=30, l=120, t=80, b=60),
            hovermode="closest"
        )
        
        # Add y-axis annotation if provided
        if y_axis_title:
            fig.add_annotation(
                x=-0.08,
                y=0 if y_range is None else (y_range[0] + y_range[1]) / 2,
                xref="paper",
                yref="y",
                text=y_axis_title,
                showarrow=False,
                font=dict(
                    size=16, 
                    family=self.config['font_family'], 
                    color="black", 
                    weight="bold"
                ),
                textangle=-90,
                align="center"
            )
    
    def _update_axes(self, fig: go.Figure, x_range: List[datetime] = None, 
                    y_range: List[float] = None, y_auto: bool = False) -> None:
        """Update the axes with common settings."""
        # Update y-axis
        y_axis_config = {
            'showgrid': True,
            'gridwidth': 1,
            'gridcolor': 'rgba(0,0,0,0.1)',
            'tickformat': ".2f",
            'tickfont': dict(
                size=self.config['axis_font_size'], 
                family=self.config['font_family'], 
                color="black", 
                weight="bold"
            ),
            'title': None
        }
        
        if y_auto:
            y_axis_config['autorange'] = True
        elif y_range:
            y_axis_config['autorange'] = False
            y_axis_config['range'] = y_range
        
        fig.update_yaxes(**y_axis_config)
        
        # Update x-axis
        x_axis_config = {
            'showgrid': True,
            'gridwidth': 1,
            'gridcolor': 'rgba(0,0,0,0.1)',
            'tickformat': "%b %d",
            'tickangle': -30,
            'tickfont': dict(
                size=self.config['axis_font_size'], 
                family=self.config['font_family'], 
                color="black", 
                weight="bold"
            ),
            'title': None
        }
        
        if x_range:
            x_axis_config['autorange'] = False
            x_axis_config['range'] = x_range
        
        fig.update_xaxes(**x_axis_config)
    
    def _add_agent_data_points(self, fig: go.Figure, df: pd.DataFrame, value_column: str,
                              color_map: Dict[str, str], max_visible: int = None) -> None:
        """Add individual agent data points to the chart."""
        if df.empty:
            return
        
        unique_agents = df['agent_name'].unique()
        max_visible = max_visible or self.config['max_visible_agents']
        
        # Calculate agent activity to determine which to show by default
        agent_counts = df['agent_name'].value_counts()
        top_agents = agent_counts.nlargest(min(max_visible, len(agent_counts))).index.tolist()
        
        logger.info(f"Showing {len(top_agents)} agents by default out of {len(unique_agents)} total agents")
        
        for agent_name in unique_agents:
            agent_data = df[df['agent_name'] == agent_name]
            
            x_values = agent_data['timestamp'].tolist()
            y_values = agent_data[value_column].tolist()
            
            # Determine visibility
            is_visible = False  # Hide all agent data points by default
            
            fig.add_trace(
                go.Scatter(
                    x=x_values,
                    y=y_values,
                    mode='markers',
                    marker=dict(
                        color=color_map.get(agent_name, 'gray'),
                        symbol='circle',
                        size=10,
                        line=dict(width=1, color='black')
                    ),
                    name=f'Agent: {agent_name} ({value_column.upper()})',
                    hovertemplate=f'Time: %{{x}}<br>{value_column.upper()}: %{{y:.2f}}<br>Agent: {agent_name}<extra></extra>',
                    visible=is_visible
                )
            )
            logger.info(f"Added {value_column} data points for agent {agent_name} with {len(x_values)} points (visible: {is_visible})")
    
    def _add_moving_average_line(self, fig: go.Figure, avg_data: pd.DataFrame, 
                                value_column: str, line_name: str, color: str, 
                                width: int = 2, hover_data: List[str] = None) -> None:
        """Add a moving average line to the chart."""
        if avg_data.empty or 'moving_avg' not in avg_data.columns:
            return
        
        # Filter out NaT values before processing - be more aggressive
        clean_data = avg_data.copy()
        
        # Remove rows with NaT timestamps more comprehensively
        clean_data = clean_data.dropna(subset=['timestamp'])
        clean_data = clean_data[clean_data['timestamp'].notna()]
        clean_data = clean_data[~clean_data['timestamp'].isnull()]
        
        # Additional check for pandas NaT specifically
        if hasattr(pd, 'NaT'):
            clean_data = clean_data[clean_data['timestamp'] != pd.NaT]
        
        # Also filter out NaN moving averages
        clean_data = clean_data.dropna(subset=['moving_avg'])
        clean_data = clean_data[clean_data['moving_avg'].notna()]
        
        if clean_data.empty:
            logger.warning("No valid timestamps found for " + str(line_name))
            return
        
        x_values = clean_data['timestamp'].tolist()
        y_values = clean_data['moving_avg'].tolist()
        
        # Create hover text without any f-strings to avoid strftime issues
        if hover_data:
            hover_text = hover_data
        else:
            hover_text = []
            for _, row in clean_data.iterrows():
                try:
                    # Convert timestamp to string safely
                    ts = row['timestamp']
                    
                    # More comprehensive NaT checking
                    if pd.isna(ts) or pd.isnull(ts) or (hasattr(pd, 'NaT') and ts is pd.NaT):
                        time_str = "Invalid Date"
                    elif hasattr(ts, 'strftime'):
                        try:
                            time_str = ts.strftime('%Y-%m-%d %H:%M:%S')
                        except (ValueError, TypeError):
                            time_str = str(ts)
                    else:
                        time_str = str(ts)
                    
                    # Build hover text using string concatenation only
                    hover_line = "Time: " + time_str + "<br>"
                    
                    # Safely format moving average value
                    try:
                        avg_val = row['moving_avg']
                        if pd.isna(avg_val) or pd.isnull(avg_val):
                            avg_str = "N/A"
                        else:
                            avg_str = "{:.2f}".format(float(avg_val))
                    except (ValueError, TypeError):
                        avg_str = "N/A"
                    
                    hover_line += "Avg " + value_column.upper() + " (7d window): " + avg_str
                    hover_text.append(hover_line)
                    
                except Exception as e:
                    logger.warning("Error formatting timestamp for hover text: " + str(e))
                    # Fallback hover text
                    hover_line = "Time: Invalid Date<br>"
                    hover_line += "Avg " + value_column.upper() + " (3d window): N/A"
                    hover_text.append(hover_line)
        
        fig.add_trace(
            go.Scatter(
                x=x_values,
                y=y_values,
                mode='lines',
                line=dict(color=color, width=width, shape='spline', smoothing=1.3),
                name=line_name,
                hovertext=hover_text,
                hoverinfo='text',
                visible=True
            )
        )
        logger.info("Added moving average line '" + str(line_name) + "' with " + str(len(x_values)) + " points")
    
    def _filter_outliers(self, df: pd.DataFrame, column: str) -> pd.DataFrame:
        """Filter outliers from the data - DISABLED: Return data unchanged."""
        # Outlier filtering disabled - return original data
        logger.info(f"Outlier filtering disabled for {column} column - returning all data")
        return df
    
    def _calculate_moving_average(self, df: pd.DataFrame, value_column: str) -> pd.DataFrame:
        """Calculate moving average for the data."""
        return self.data_processor.calculate_moving_average(df, value_column)
    
    def _save_chart(self, fig: go.Figure, html_filename: str, png_filename: str = None) -> None:
        """Save the chart to HTML and optionally PNG."""
        try:
            fig.write_html(html_filename, include_plotlyjs='cdn', full_html=False)
            logger.info(f"Chart saved to {html_filename}")
            
            if png_filename:
                try:
                    fig.write_image(png_filename)
                    logger.info(f"Chart also saved to {png_filename}")
                except Exception as e:
                    logger.error(f"Error saving PNG image: {e}")
                    logger.info(f"Chart saved to {html_filename} only")
        except Exception as e:
            logger.error(f"Error saving chart: {e}")
    
    def generate_visualization(self, df: pd.DataFrame, **kwargs) -> Tuple[go.Figure, Optional[str]]:
        """Generate the complete visualization including chart and CSV export."""
        if df.empty:
            logger.info("No data available for visualization.")
            fig = self._create_empty_chart("No data available")
            return fig, None
        
        # Create the chart
        fig = self.create_chart(df, **kwargs)
        
        # Save to CSV
        csv_filename = kwargs.get('csv_filename')
        if csv_filename:
            csv_path = self.data_processor.save_to_csv(df, csv_filename)
        else:
            csv_path = None
        
        return fig, csv_path
    
    def _create_empty_chart(self, message: str) -> go.Figure:
        """Create an empty chart with a message."""
        fig = go.Figure()
        fig.add_annotation(
            x=0.5, y=0.5,
            text=message,
            font=dict(size=20),
            showarrow=False
        )
        fig.update_layout(
            xaxis=dict(showgrid=False, zeroline=False, showticklabels=False),
            yaxis=dict(showgrid=False, zeroline=False, showticklabels=False)
        )
        return fig
    
    def _get_color_map(self, agents: List[str]) -> Dict[str, str]:
        """Generate a color map for agents."""
        colors = px.colors.qualitative.Plotly[:len(agents)]
        return {agent: colors[i % len(colors)] for i, agent in enumerate(agents)}