src/visualizations.py

import pandas as pd
import numpy as np
import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import folium
from streamlit_folium import folium_static
import streamlit as st

def create_price_distribution_chart(data, predicted_price=None):
    """
    Crée un histogramme de la distribution des prix avec la prédiction
    """
    if data is None or 'price' not in data.columns:
        return None
    
    fig = px.histogram(
        data, 
        x='price', 
        nbins=50,
        title="     Distribution des Prix Immobiliers",
        labels={'price': 'Prix ($)', 'count': 'Nombre de propriétés'},
        color_discrete_sequence=['#3498db']
    )
    
    # Ajouter une ligne verticale pour la prédiction
    if predicted_price is not None:
        fig.add_vline(
            x=predicted_price, 
            line_dash="dash", 
            line_color="red",
            annotation_text=f"Votre prédiction: ${predicted_price:,.0f}",
            annotation_position="top right"
        )
    
    fig.update_layout(
        template="plotly_dark",
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font=dict(color='#D4DCFF'),
        title_font_size=20,
        showlegend=False
    )
    
    return fig

def create_feature_impact_chart(data, feature_name, predicted_price=None):
    """
    Crée un graphique montrant l'impact d'une caractéristique sur le prix
    """
    if data is None or feature_name not in data.columns or 'price' not in data.columns:
        return None
    
    # Calculer la moyenne des prix par valeur de la caractéristique
    feature_impact = data.groupby(feature_name)['price'].mean().reset_index()
    
    fig = px.line(
        feature_impact,
        x=feature_name,
        y='price',
        title=f"     Impact de {feature_name} sur le Prix",
        labels={'price': 'Prix Moyen ($)', feature_name: feature_name.title()},
        markers=True
    )
    
    # Ajouter une ligne horizontale pour la prédiction
    if predicted_price is not None:
        fig.add_hline(
            y=predicted_price,
            line_dash="dash",
            line_color="red",
            annotation_text=f"Votre prédiction: ${predicted_price:,.0f}",
            annotation_position="top right"
        )
    
    fig.update_layout(
        template="plotly_dark",
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font=dict(color='#D4DCFF'),
        title_font_size=16,
        title=dict(
            x=0.5,
            xanchor='center',
            y=0.95,
            yanchor='top',
            pad=dict(t=20, b=30, l=20, r=20)
        )
    )
    
    return fig

def create_price_vs_area_chart(data, predicted_price=None, predicted_area=None):
    """
    Crée un graphique de dispersion prix vs surface
    """
    if data is None or 'sqft_living' not in data.columns or 'price' not in data.columns:
        return None
    
    fig = px.scatter(
        data,
        x='sqft_living',
        y='price',
        title="     Prix vs Surface Habitable",
        labels={'sqft_living': 'Surface Habitable (pieds²)', 'price': 'Prix ($)'},
        opacity=0.6,
        color_discrete_sequence=['#3498db']
    )
    
    # Ajouter le point de prédiction
    if predicted_price is not None and predicted_area is not None:
        fig.add_scatter(
            x=[predicted_area],
            y=[predicted_price],
            mode='markers',
            marker=dict(size=15, color='red', symbol='star'),
            name='Votre prédiction',
            showlegend=True
        )
    
    fig.update_layout(
        template="plotly_dark",
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font=dict(color='#D4DCFF'),
        title_font_size=16
    )
    
    return fig

def create_correlation_heatmap(data):
    """
    Crée une heatmap de corrélation entre les caractéristiques
    """
    if data is None:
        return None
    
    # Sélectionner les colonnes numériques
    numeric_cols = data.select_dtypes(include=[np.number]).columns
    if len(numeric_cols) < 2:
        return None
    
    correlation_matrix = data[numeric_cols].corr()
    
    fig = px.imshow(
        correlation_matrix,
        title="     Matrice de Corrélation des Caractéristiques",
        color_continuous_scale='RdBu',
        aspect="auto"
    )
    
    fig.update_layout(
        template="plotly_dark",
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font=dict(color='#D4DCFF'),
        title_font_size=16
    )
    
    return fig

def create_interactive_map(data, lat=None, long=None, predicted_price=None):
    """
    Crée une carte interactive avec les propriétés et la prédiction
    """
    if data is None or 'lat' not in data.columns or 'long' not in data.columns:
        return None
    
    # Calculer le centre de la carte
    center_lat = data['lat'].mean() if lat is None else lat
    center_long = data['long'].mean() if long is None else long
    
    # Créer la carte
    m = folium.Map(
        location=[center_lat, center_long],
        zoom_start=10,
        tiles='cartodbpositron'
    )
    
    # Ajouter les propriétés existantes (échantillon pour éviter la surcharge)
    sample_data = data.sample(min(100, len(data)))
    
    for idx, row in sample_data.iterrows():
        folium.CircleMarker(
            location=[row['lat'], row['long']],
            radius=3,
            popup=f"Prix: ${row['price']:,.0f}<br>Surface: {row['sqft_living']} pieds²",
            color='blue',
            fill=True,
            fillOpacity=0.7
        ).add_to(m)
    
    # Ajouter la prédiction si disponible
    if lat is not None and long is not None and predicted_price is not None:
        folium.Marker(
            location=[lat, long],
            popup=f"Votre prédiction: ${predicted_price:,.0f}",
            icon=folium.Icon(color='red', icon='star')
        ).add_to(m)
    
    return m

def create_price_by_zipcode_chart(data, predicted_price=None, predicted_zipcode=None):
    """
    Crée un graphique des prix moyens par code postal
    """
    if data is None or 'zipcode' not in data.columns or 'price' not in data.columns:
        return None
    
    # Calculer le prix moyen par code postal
    zipcode_prices = data.groupby('zipcode')['price'].mean().reset_index()
    zipcode_prices = zipcode_prices.sort_values('price', ascending=False).head(20)
    
    fig = px.bar(
        zipcode_prices,
        x='zipcode',
        y='price',
        title="     Prix Moyen par Code Postal (Top 20)",
        labels={'price': 'Prix Moyen ($)', 'zipcode': 'Code Postal'},
        color='price',
        color_continuous_scale='viridis'
    )
    
    # Ajouter une ligne horizontale pour la prédiction
    if predicted_price is not None:
        fig.add_hline(
            y=predicted_price,
            line_dash="dash",
            line_color="red",
            annotation_text=f"Votre prédiction: ${predicted_price:,.0f}",
            annotation_position="top right"
        )
    
    fig.update_layout(
        template="plotly_dark",
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font=dict(color='#D4DCFF'),
        title_font_size=16,
        xaxis_tickangle=-45,
        title=dict(
            x=0.5,
            xanchor='center',
            y=0.95,
            yanchor='top',
            pad=dict(t=20, b=30, l=20, r=20)
        )
    )
    
    return fig

def create_comparison_dashboard(data, predicted_price, input_features):
    """
    Crée un tableau de bord de comparaison
    """
    if data is None:
        return None
    
    # Calculer les statistiques de comparaison
    stats = {
        'Prix moyen du marché': f"${data['price'].mean():,.0f}",
        'Prix médian du marché': f"${data['price'].median():,.0f}",
        'Votre prédiction': f"${predicted_price:,.0f}",
        'Différence avec la moyenne': f"${predicted_price - data['price'].mean():,.0f}",
        'Pourcentage vs moyenne': f"{((predicted_price / data['price'].mean()) - 1) * 100:.1f}%"
    }
    
    # Créer un graphique de comparaison
    comparison_data = pd.DataFrame({
        'Métrique': ['Prix Moyen', 'Prix Médian', 'Votre Prédiction'],
        'Prix': [data['price'].mean(), data['price'].median(), predicted_price]
    })
    
    fig = px.bar(
        comparison_data,
        x='Métrique',
        y='Prix',
        title="     Comparaison avec le Marché",
        color='Métrique',
        color_discrete_map={
            'Prix Moyen': '#3498db',
            'Prix Médian': '#2ecc71',
            'Votre Prédiction': '#e74c3c'
        }
    )
    
    fig.update_layout(
        template="plotly_dark",
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font=dict(color='#D4DCFF'),
        title_font_size=16,
        showlegend=False,
        title=dict(
            x=0.5,
            xanchor='center',
            y=0.95,
            yanchor='top',
            pad=dict(t=20, b=30, l=20, r=20)
        )
    )
    
    return fig, stats