import os
import pandas as pd
import numpy as np
import geopandas as gpd
import plotly.express as px
from dash import Dash, dcc, html, Input, Output, callback
import dash_bootstrap_components as dbc

external_stylesheets = ['https://codepen.io/chriddyp/pen/bWLwgP.css']

app = Dash(__name__, external_stylesheets=[dbc.themes.SLATE])
server = app.server  # Required for deployment

# Load data using relative paths
flood_risk = gpd.read_file("data/final_flood_risk_scores.geojson")
flood_risk['Census Tract'] = flood_risk['Census Tract'].astype(int)

ami = pd.read_csv("data/acs_2023_ami_cleaned.csv")
age = pd.read_csv("data/acs_2023_age_cleaned.csv")
tenure = pd.read_csv("data/acs_2023_tenure_cleaned.csv")
race = pd.read_csv("data/race_percentages.csv")

# Adjust race dataframe for display
race[['NH White','NH Black or African American','NH Asian','Hispanic or Latino','NH Other']] = race[['NH White','NH Black or African American','NH Asian','Hispanic or Latino','NH Other']] * 100
race = race.rename(columns={'NH White': 'White', 'NH Black or African American': 'Black or African American','NH Asian' : 'Asian', 'NH Other': 'Other'})
race = race.set_index("geoid")
race.rename(index={'total': '111'}, inplace=True)
race.index = race.index.astype(int)
race.rename(index={111: 'San Francisco'}, inplace=True)


def normalize_scores(series):
    """Takes a series and normalizes values to be on a 0-100 range"""
    max_val = series.max()
    min_val = series.min()
    scores = ((series - min_val) * 100) / (max_val - min_val)
    return scores.round(1)


def race_plot(tract_id: str):
    rows = race.loc[['San Francisco', tract_id], race.columns[0:5]]
    race_fig = px.bar(
        rows.T,
        y=rows.index,
        x=rows.columns,
        barmode="group",
        labels={'geoid': 'Geographic Area', 'index': 'Race or Ethnicity', 'value': 'Percent of Total Population'},
        color_discrete_sequence=['lightgray', 'black']
    )
    race_fig.update_layout(xaxis={'categoryorder': 'total descending'})
    race_fig.update_traces(
        hovertemplate="%{x}<br><b>%{y:,.0f}%<br></b><extra></extra>"
    )
    race_fig.update_layout({'plot_bgcolor': 'rgba(0,0,0,0)'})
    return race_fig


def baseline_race_plot():
    rows = race.loc[['San Francisco'], race.columns[0:5]]
    race_fig = px.bar(
        rows.T,
        y=rows.index,
        x=rows.columns,
        barmode="group",
        labels={'geoid': 'Geographic Area', 'index': 'Race or Ethnicity', 'value': 'Percent of Total Population'},
        color_discrete_sequence=['lightgray', 'black']
    )
    race_fig.update_layout(xaxis={'categoryorder': 'total descending'})
    race_fig.update_traces(
        hovertemplate="%{x}<br><b>%{y:,.0f}%<br></b><extra></extra>"
    )
    race_fig.update_layout({'plot_bgcolor': 'rgba(0,0,0,0)'})
    return race_fig


def plot_data(df: pd.DataFrame, tract_id: str, column: str, x_label: str):
    df = df.copy()
    df['select'] = 0
    df['opacity'] = 0.5
    df.loc[df['geoid'] == tract_id, 'select'] = 1
    df.loc[df['geoid'] == tract_id, 'opacity'] = 1
    fig = px.strip(
        df,
        x=column,
        color='select',
        stripmode='overlay',
        color_discrete_map={1: 'black', 0: 'darkgray'},
        labels={column: x_label},
        hover_name='geoid',
        hover_data={column: True, 'select': False}
    )
    fig.update_layout(showlegend=False)
    fig.update_layout({'plot_bgcolor': 'rgba(0,0,0,0)'})
    fig.update_layout(margin=dict(l=0, r=0, t=0, b=0))
    fig.update_layout(height=250)
    fig.update_traces({'marker': {'size': 15}})
    fig.update_xaxes(gridcolor='#f2f2f2')
    fig.update_traces(
        hovertemplate=f"Census Tract: {tract_id}<br>{x_label}: <b>%{{x:,.0f}}<br></b><extra></extra>"
    )
    fig.data[0].marker.opacity = 0.2
    fig.data[1].marker.opacity = 1.0
    return fig


def baseline_plot(df: pd.DataFrame, column: str, x_label: str):
    df = df.copy()
    df['select'] = 0
    df['opacity'] = 0.5
    fig = px.strip(
        df,
        x=column,
        color_discrete_sequence=['darkgray'],
        stripmode='overlay',
        labels={column: x_label},
        hover_name='geoid',
        hover_data={column: True, 'select': False},
        custom_data=['geoid']
    )
    fig.update_layout(showlegend=False)
    fig.update_layout({'plot_bgcolor': 'rgba(0,0,0,0)'})
    fig.update_layout(margin=dict(l=0, r=0, t=0, b=0))
    fig.update_layout(height=250)
    fig.update_traces({'marker': {'size': 15}})
    fig.update_xaxes(gridcolor='#f2f2f2')
    fig.update_traces(marker=dict(opacity=0.2))
    fig.update_traces(
        hovertemplate="Census Tract: %{customdata[0]}<br>" + x_label + ": <b>%{x:,.0f}<br></b><extra></extra>"
    )
    return fig


def update_figure(value1, value2, value3):
    updated_flood = flood_risk.copy()
    total_value = value1 + value2 + value3
    factor1 = value1 / total_value
    factor2 = value2 / total_value
    factor3 = value3 / total_value
    updated_flood["Weighted flood risk score"] = (
        updated_flood["Floodplain coverage score"] * factor1 +
        updated_flood["Drain quality score"] * factor2 +
        updated_flood["Green infrastructure score"] * factor3
    )
    updated_flood["Weighted flood risk score"] = normalize_scores(updated_flood["Weighted flood risk score"])

    sf_lat = 37.7749
    sf_lon = -122.4194
    updated_flood = updated_flood.set_index('Census Tract')
    updated_flood = updated_flood.to_crs("EPSG:4326")
    fig = px.choropleth_map(
        updated_flood,
        geojson=updated_flood.geometry,
        locations=updated_flood.index,
        color="Weighted flood risk score",
        center={"lat": sf_lat, "lon": sf_lon},
        map_style="carto-positron",
        zoom=10,
        color_continuous_scale='blues'
    )
    fig.update_layout(
        transition_duration=200,
        uirevision="constant",
        clickmode='event+select',
        width=800,
        height=400,
        margin={"r": 0, "t": 0, "l": 0, "b": 0},
        coloraxis_colorbar=dict(orientation='h', len=0.8),
        coloraxis_colorbar_title_font=dict(family='sans-serif')
    )
    fig.update_traces(
        marker_line_width=0.5,
        marker_opacity=0.75,
        marker_line_color='white'
    )
    fig.update_traces(
        selected=dict(marker=dict(opacity=1)),
        unselected=dict(marker=dict(opacity=0.5))
    )
    fig.update_traces(
        hovertemplate="Census Tract: %{location}<br>Weighted flood risk score: <b>%{z:,.0f}</b><extra></extra>"
    )
    return fig


def update_graphs(clicked_data):
    if clicked_data is not None:
        tract_id = clicked_data["points"][0]["location"]
        ami_fig = plot_data(ami, tract_id, "median_income", 'Median Income')
        tenure_fig = plot_data(tenure, tract_id, "pct_renter", 'Percent Renter Households')
        under5_fig = plot_data(age, tract_id, 'Under 5 Years', 'Percent Residents Under 5 Years')
        over65_fig = plot_data(age, tract_id, '65 Years and Over', 'Percent Residents 65 Years and Older')
        race_fig = race_plot(tract_id)
        return ami_fig, tenure_fig, under5_fig, over65_fig, race_fig
    else:
        ami_fig = baseline_plot(ami, "median_income", "Median Income")
        tenure_fig = baseline_plot(tenure, "pct_renter", 'Percent Renter Households')
        under5_fig = baseline_plot(age, 'Under 5 Years', 'Percent Residents Under 5 Years')
        over65_fig = baseline_plot(age, '65 Years and Over', 'Percent Residents 65 Years and Older')
        race_fig = baseline_race_plot()
        return ami_fig, tenure_fig, under5_fig, over65_fig, race_fig


# Set initial state
initial_ami_fig, initial_tenure_fig, initial_under5_fig, initial_over65_fig, initial_race_fig = update_graphs(None)
initial_flood_map = update_figure(2, 2, 2)

# Build the app layout
app.layout = html.Div([
    html.Div(
        children="Analyzing Flood Risk in San Francisco",
        style={'fontFamily': 'Arial', 'fontSize': '40px', "font-weight": "bold", 'textAlign': 'center', 'paddingTop': '30px'}
    ),
    html.Div([
        # Sliders
        html.Div([
            html.Div([
                html.H3("Select a floodplain coverage weight:", style={'fontFamily': 'Arial', 'fontSize': '18px', "font-weight": "normal", 'letterSpacing': '.05px'}),
                dcc.Slider(0, 4, step=None, marks={0: 'low priority', 1: '', 2: 'moderate priority', 3: '', 4: 'high priority'}, value=2, id='floodplain_slider')
            ]),
            html.Div([
                html.H3("Select a drain quality weight:", style={'fontFamily': 'Arial', 'fontSize': '18px', 'letterSpacing': '.05px'}),
                dcc.Slider(0, 4, step=None, marks={0: 'low priority', 1: '', 2: 'moderate priority', 3: '', 4: 'high priority'}, value=2, id='drain_slider')
            ]),
            html.Div([
                html.H3("Select a green infrastructure weight:", style={'fontFamily': 'Arial', 'fontSize': '18px', 'letterSpacing': '.05px'}),
                dcc.Slider(0, 4, step=None, marks={0: 'low priority', 1: '', 2: 'moderate priority', 3: '', 4: 'high priority'}, value=2, id='gi_slider')
            ]),
        ], style={'display': 'flex', 'flexDirection': 'column', 'flex': 1, 'marginTop': '20px', 'paddingRight': '40px', 'padding': '3%', 'justifyContent': 'center', "gap": "25px"}),
        html.Div(id='slider-output-container'),
        # Flood risk score map
        html.Div(dcc.Graph(id='flood-risk-graph', figure=initial_flood_map), style={'padding': '3%', 'flex': 1})
    ], style={'display': 'flex', 'flexDirection': 'row'}),

    # Census tract name display
    html.Div([
        html.Span("Census Tract:", style={'fontFamily': 'Arial', 'fontSize': '18px', 'marginRight': '5px'}),
        html.Span(id='tract_id', style={'fontFamily': 'Arial', 'fontSize': '18px', 'font-weight': 'bold'})
    ], style={'display': 'flex', 'flexDirection': 'row', 'padding': '3%', 'justify-content': 'center'}),

    # AMI and tenure charts
    html.Div([
        html.Div(dcc.Graph(id='ami_fig', figure=initial_ami_fig), style={'padding': '3%', 'flex': 1}),
        html.Div(dcc.Graph(id='tenure_fig', figure=initial_tenure_fig), style={'padding': '3%', 'flex': 1}),
    ], style={'display': 'flex', 'flexDirection': 'row'}),

    # Age charts
    html.Div([
        html.Div(dcc.Graph(id='under5_fig', figure=initial_under5_fig), style={'padding': '3%', 'flex': 1}),
        html.Div(dcc.Graph(id='over65_fig', figure=initial_over65_fig), style={'padding': '3%', 'flex': 1}),
    ], style={'display': 'flex', 'flexDirection': 'row'}),

    # Race bar chart
    html.Div([
        html.Div(dcc.Graph(id='race_fig', figure=initial_race_fig), style={'padding': '3%', 'flex': 1}),
    ], style={'display': 'flex', 'flexDirection': 'row', 'paddingLeft': '20%', 'paddingRight': '20%'}),
])


@callback(
    [Output('flood-risk-graph', 'figure'),
     Output('ami_fig', 'figure'),
     Output('tenure_fig', 'figure'),
     Output('under5_fig', 'figure'),
     Output('over65_fig', 'figure'),
     Output('race_fig', 'figure'),
     Output('tract_id', 'children')],
    [Input('floodplain_slider', 'value'),
     Input('drain_slider', 'value'),
     Input('gi_slider', 'value'),
     Input('flood-risk-graph', 'clickData')]
)
def update_output(value1, value2, value3, clickData):
    flood_risk_fig = update_figure(value1, value2, value3)
    ami_fig, tenure_fig, under5_fig, over65_fig, race_fig = update_graphs(clickData)
    if clickData is not None:
        tract_id = clickData["points"][0]["location"]
    else:
        tract_id = 'Select a census tract on the map'
    return flood_risk_fig, ami_fig, tenure_fig, under5_fig, over65_fig, race_fig, tract_id


if __name__ == '__main__':
    app.run(debug=False, host="0.0.0.0", port=7860)