visualizations/segment_ranking.py

# visualizations/segment_ranking.py

import plotly.graph_objects as go
import plotly.express as px
from plotly.subplots import make_subplots
import pandas as pd
from metrics.metric_registry import METRIC_FUNCTIONS
from analytics.performance_analysis import generate_metric_view


def calculate_segment_risk_score(
    df,
    metric_name,
    category
):
    """
    Calculate dollar-based risk scores for each segment in a category.
    
    Args:
        df: Master dataframe
        metric_name: Metric name for risk calculation
        category: Segmentation category
    
    Returns:
        DataFrame with segment and risk score (dollar-based %)
    """
    
    result = generate_metric_view(
        df=df,
        metric_name=metric_name,
        group_col=category
    )
    
    rate_col = [
        col for col in result.columns
        if "rate" in col.lower()
    ][0]
    
    # Calculate dollar-based risk per segment
    # Risk = (Total Bad Balance) / (Total Balance) * 100
    segment_risk = (
        result.groupby(category)
        .agg({
            rate_col: "mean",
            "total_accounts": "sum",
            "total_balance": "sum"
        })
        .reset_index()
    )
    
    segment_risk = segment_risk.rename(
        columns={
            category: "Segment",
            rate_col: "Risk_Score"
        }
    )
    
    return segment_risk


def generate_segment_risk_heatmap(
    df,
    metrics=None,
    categories=None
):
    """
    Generate heatmap showing risk scores across segments and metrics.
    
    Args:
        df: Master dataframe
        metrics: List of metrics to evaluate
        categories: List of categories to analyze
    
    Returns:
        Plotly figure with heatmap
    """
    
    if metrics is None:
        metrics = ["30+@3", "30+@6", "60+@6", "Yr1 NCL"]
    
    if categories is None:
        categories = [
            "fico_band",
            "sourcing_channel",
            "city_tier",
            "occupation_type"
        ]
    
    # Prepare data for heatmap
    heatmap_data = {}
    all_segments = {}
    
    for metric in metrics:
        
        metric_scores = {}
        
        for category in categories:
            
            try:
                segment_risk = calculate_segment_risk_score(
                    df=df,
                    metric_name=metric,
                    category=category
                )
                
                for _, row in segment_risk.iterrows():
                    segment_key = f"{category}_{row['Segment']}"
                    metric_scores[segment_key] = row["Risk_Score"]
                    all_segments[segment_key] = f"{category.replace('_', ' ').title()}: {row['Segment']}"
            
            except Exception as e:
                print(f"Error processing {metric} x {category}: {e}")
        
        heatmap_data[metric] = metric_scores
    
    # Create DataFrame for heatmap
    heatmap_df = pd.DataFrame(heatmap_data)
    heatmap_df = heatmap_df.fillna(0)
    
    # Sort by average risk
    heatmap_df["avg_risk"] = heatmap_df.mean(axis=1)
    heatmap_df = heatmap_df.sort_values("avg_risk", ascending=False)
    heatmap_df = heatmap_df.drop("avg_risk", axis=1)
    
    # Create heatmap
    fig = go.Figure(
        data=go.Heatmap(
            z=heatmap_df.values,
            x=heatmap_df.columns,
            y=[all_segments.get(idx, idx) for idx in heatmap_df.index],
            colorscale="RdYlGn_r",
            hovertemplate=(
                "<b>Segment: %{y}</b><br>" +
                "<b>Metric: %{x}</b><br>" +
                "Risk Score: %{z:.2f}%<br>" +
                "<extra></extra>"
            ),
            text=[[f"{val:.2f}%" for val in row] for row in heatmap_df.values],
            texttemplate="%{text}",
            textfont={"size": 10},
            colorbar=dict(
                title="Risk Score<br>(%)"
            )
        )
    )
    
    fig.update_layout(
        title="Segment Risk Heatmap Across Delinquency Metrics",
        xaxis_title="Delinquency Metrics",
        yaxis_title="Segments",
        height=max(400, len(heatmap_df) * 25),
        template="plotly_white",
        hovermode="closest"
    )
    
    return fig


def generate_segment_risk_ranking(
    df,
    metric_name,
    category,
    top_n=10
):
    """
    Generate bar chart ranking segments by risk within a category.
    
    Args:
        df: Master dataframe
        metric_name: Metric name for risk calculation
        category: Segmentation category
        top_n: Number of top risk segments to display
    
    Returns:
        Plotly bar chart figure
    """
    
    segment_risk = calculate_segment_risk_score(
        df=df,
        metric_name=metric_name,
        category=category
    )
    
    # Sort by risk score descending
    segment_risk = segment_risk.sort_values(
        "Risk_Score",
        ascending=True
    ).tail(top_n)
    
    # Color code by risk level
    colors = ["#d62728" if score > 10 else "#ff7f0e" if score > 5 else "#2ca02c"
              for score in segment_risk["Risk_Score"]]
    
    fig = go.Figure(
        data=go.Bar(
            y=segment_risk["Segment"],
            x=segment_risk["Risk_Score"],
            orientation="h",
            marker=dict(
                color=colors,
                line=dict(color="white", width=1)
            ),
            text=segment_risk["Risk_Score"],
            texttemplate="%{text:.2f}%",
            textposition="outside",
            hovertemplate=(
                "<b>Segment: %{y}</b><br>" +
                "Risk Score: %{x:.2f}%<br>" +
                "Accounts: %{customdata[0]}<br>" +
                "Balance: %{customdata[1]:,.0f}<br>" +
                "<extra></extra>"
            ),
            customdata=segment_risk[["total_accounts", "total_balance"]].values
        )
    )
    
    fig.update_layout(
        title=f"Top {top_n} High-Risk Segments: {metric_name} by {category.replace('_', ' ').title()}",
        xaxis_title="Risk Score (%)",
        yaxis_title=category.replace('_', ' ').title(),
        height=400 + (top_n * 15),
        template="plotly_white",
        hovermode="closest"
    )
    
    fig.update_xaxes(
        showgrid=True,
        gridwidth=1,
        gridcolor="lightgray"
    )
    
    return fig


def generate_multi_category_risk_comparison(
    df,
    metric_name
):
    """
    Compare risk across all categories for a single metric.
    
    Args:
        df: Master dataframe
        metric_name: Metric name for risk calculation
    
    Returns:
        Plotly figure with subplots (one per category)
    """
    
    categories = [
        "fico_band",
        "sourcing_channel",
        "city_tier",
        "occupation_type"
    ]
    
    # Create subplots
    fig = make_subplots(
        rows=2,
        cols=2,
        subplot_titles=[cat.replace('_', ' ').title() for cat in categories],
        specs=[
            [{"type": "bar"}, {"type": "bar"}],
            [{"type": "bar"}, {"type": "bar"}]
        ]
    )
    
    positions = [
        (1, 1),
        (1, 2),
        (2, 1),
        (2, 2)
    ]
    
    max_segments = 0
    
    for category, (row, col) in zip(categories, positions):
        
        try:
            segment_risk = calculate_segment_risk_score(
                df=df,
                metric_name=metric_name,
                category=category
            )
            
            # Sort and take top 5
            segment_risk = segment_risk.sort_values(
                "Risk_Score",
                ascending=True
            ).tail(5)
            
            max_segments = max(max_segments, len(segment_risk))
            
            fig.add_trace(
                go.Bar(
                    y=segment_risk["Segment"],
                    x=segment_risk["Risk_Score"],
                    orientation="h",
                    name=category,
                    showlegend=False,
                    marker=dict(
                        color=segment_risk["Risk_Score"],
                        colorscale="Reds",
                        showscale=False
                    ),
                    text=segment_risk["Risk_Score"],
                    texttemplate="%{text:.2f}%",
                    textposition="outside",
                    hovertemplate=(
                        "<b>%{y}</b><br>" +
                        "Risk Score: %{x:.2f}%<br>" +
                        "<extra></extra>"
                    )
                ),
                row=row,
                col=col
            )
            
            fig.update_xaxes(
                title_text="Risk Score (%)",
                row=row,
                col=col
            )
        
        except Exception as e:
            print(f"Error processing category {category}: {e}")
    
    fig.update_layout(
        title_text=f"High-Risk Segments Across Categories: {metric_name}",
        height=800,
        template="plotly_white",
        hovermode="closest"
    )
    
    return fig


def calculate_portfolio_risk_summary(
    df,
    metrics=None
):
    """
    Calculate overall portfolio risk summary across metrics and categories.
    
    Args:
        df: Master dataframe
        metrics: List of metrics to evaluate
    
    Returns:
        DataFrame with portfolio risk summary
    """
    
    if metrics is None:
        metrics = ["30+@3", "30+@6", "60+@6", "Yr1 NCL"]
    
    summary_data = []
    
    categories = [
        "fico_band",
        "sourcing_channel",
        "city_tier",
        "occupation_type"
    ]
    
    for metric in metrics:
        for category in categories:
            try:
                segment_risk = calculate_segment_risk_score(
                    df=df,
                    metric_name=metric,
                    category=category
                )
                
                avg_risk = segment_risk["Risk_Score"].mean()
                max_risk = segment_risk["Risk_Score"].max()
                high_risk_count = len(segment_risk[segment_risk["Risk_Score"] > 10])
                
                summary_data.append({
                    "Metric": metric,
                    "Category": category.replace('_', ' ').title(),
                    "Avg_Risk": avg_risk,
                    "Max_Risk": max_risk,
                    "High_Risk_Segments": high_risk_count
                })
            
            except Exception as e:
                print(f"Error calculating summary for {metric} x {category}: {e}")
    
    summary_df = pd.DataFrame(summary_data)
    
    return summary_df