Py.Cafe

# -*- coding: utf-8 -*-
"""
Created on Fri May 16 21:01:26 2025

@author: Marie-Anne Melis
"""

import pandas as pd
import dash
from dash import dcc, html
import plotly.graph_objects as go
from dash.dependencies import Input, Output
import dash_bootstrap_components as dbc
from dash_bootstrap_templates import load_figure_template
from plotly.subplots import make_subplots


colors_class = ['rgba(184,0,127,.2)', 
                'rgba(184,0,127,.4)',
                'rgba(184,0,127,.7)',
                'rgba(184,0,127,1)']


#df_raw = pd.read_csv('nation.csv')
#df = df_raw[['Federal ID','Hazard Potential Classification','Primary Owner Type','Primary Purpose', 'EAP Prepared']].reset_index()
#df.to_csv('nation-dams.csv')
df=pd.read_csv('nation-dams.csv')


# stylesheet with the .dbc class to style  dcc, DataTable and AG Grid components with a Bootstrap theme
dbc_css = "https://cdn.jsdelivr.net/gh/AnnMarieW/dash-bootstrap-templates/dbc.min.css"
# if using the vizro theme
vizro_bootstrap = "https://cdn.jsdelivr.net/gh/mckinsey/vizro@main/vizro-core/src/vizro/static/css/vizro-bootstrap.min.css?v=2"

# default dark mode
#pio.templates.default = "vizro_dark"
load_figure_template(["vizro", "vizro_dark"])


#subplot reference https://fronkan.hashnode.dev/different-barmodes-simultaneously-in-plotly-subplots-python


#df6 == PURPOSE

dff6 = df.groupby(['Hazard Potential Classification','Primary Purpose', 'EAP Prepared'])['Federal ID'].count().reset_index()
dff6.rename(columns={"Federal ID": "Number"}, inplace=True)
# Calculate percentage within each hazard group
dff6["Percentage Hazard"] = round(dff6["Number"] / dff6.groupby("Hazard Potential Classification")["Number"].transform("sum") * 100,1)
dff6["Percentage Hazard Purpose"] = round(dff6["Number"] / dff6.groupby(["Hazard Potential Classification","Primary Purpose"])["Number"].transform("sum") * 100,1)

#df2 = OWNER

dff2 = df.groupby(['Hazard Potential Classification','Primary Owner Type', 'EAP Prepared'])['Federal ID'].count().reset_index()
dff2.rename(columns={"Federal ID": "Number"}, inplace=True)
# Calculate percentage within each hazard group
dff2["Percentage Hazard"] = round(dff2["Number"] / dff2.groupby("Hazard Potential Classification")["Number"].transform("sum") * 100,1)
dff2["Percentage Hazard Owner"] = round(dff2["Number"] / dff2.groupby(["Hazard Potential Classification","Primary Owner Type"])["Number"].transform("sum") * 100,1)


#   choose between owners and purpose view
radioitems = html.Div(
    [
        dbc.Label("Zoom in on:"),
        dbc.RadioItems(
            options=[
                {"label": "Owner dam", "value": 'Owner'},
                {"label": "Purpose dam", "value": "Purpose"},
            ],
            value="Owner",
            id="radioitems-input",
            inline=True 
        ),
    ]
)





def create_summary(df):
    
    #FUNCTION CREATES THE GENERAL OVERVIEW OF HAZARD CLASSIFICATION AND EAP 
    
    
    # Aggregate
    dff4 = df.groupby(['Hazard Potential Classification', 'EAP Prepared'])['Federal ID'].count().reset_index()
    dff4.rename(columns={"Federal ID": "Number"}, inplace=True)

    # Percentages within group
    dff4["Percentage"] = round(
        dff4["Number"] / dff4.groupby("Hazard Potential Classification")["Number"].transform("sum") * 100, 1
    )

    # Desired order and colors
    eap_statuses = ['No', 'Not Required', 'Yes']
    colors = ['rgba(0, 61, 84,1)', 
              'rgba(56, 111, 152, 1)', 
              'rgba(188, 212, 230, 1)']

    # Set correct order for y-axis
    hazard_order = ["Undetermined", "Low", "Significant", "High"]
    dff4["Hazard Potential Classification"] = pd.Categorical(
        dff4["Hazard Potential Classification"],
        categories=hazard_order,
        ordered=True
    )
    dff4 = dff4.sort_values("Hazard Potential Classification")
    
    #dfs contains number of dambs per classification
    dfs = df['Hazard Potential Classification'].value_counts().reset_index()
    dfs["Hazard Potential Classification"] = pd.Categorical(
        dfs["Hazard Potential Classification"],
        categories=hazard_order,
        ordered=True
    )
    dfs = dfs.sort_values("Hazard Potential Classification")
    
    
    # Creating two subplots
    fig = make_subplots(rows=1, cols=2,specs=[[{}, {}]], shared_xaxes=True,
                    shared_yaxes=False, vertical_spacing=0.01,
                    subplot_titles=("USA: number of dams per Hazard Potential Classification", "Emergency Action Plan prepared?"))
    
    
    

    # Build traces for subplot 2, the percentages
    #traces = []
    for status, color in zip(eap_statuses, colors):
        subset = dff4[dff4["EAP Prepared"] == status]

        fig.add_trace(
            go.Bar(
                y=subset["Hazard Potential Classification"],
                x=subset["Percentage"],
                name=status,
                orientation="h",
                offsetgroup=1,
                marker_color=color,
                text=subset["Percentage"].astype(str) + '%',
                textposition="inside",
                insidetextanchor="middle",

            ),
            row=1,
            col=2,
        ),
       
      # Subplot1 - Overview
    fig.add_trace(
          go.Bar(
              
              x=dfs['count'],
              y=dfs['Hazard Potential Classification'],
              orientation="h",
              marker_color=colors_class,
              offsetgroup=0,
              showlegend=False,
              text=dfs['count'],
                textposition="outside",
                insidetextanchor="end",
              ),
          row=1,
          col=1,
          
      )      

    #add a warning rectangle for class High , no and not required
    fig.add_shape(type="rect",
        
        x0=0, y0=2.6, x1=23.7, y1=3.4,
        line=dict(
            color=colors_class[3],
            width=1,
        ),
        fillcolor=colors_class[1],
          row=1,
          col=2,
    )
    
    #add a warning rectangle for class Significant , no and not required
    fig.add_shape(type="rect",
        
        x0=0, y0=1.6, x1=34.3, y1=2.4,
        line=dict(
            color=colors_class[3],
            width=1,
        ),
        fillcolor=colors_class[1],
          row=1,
          col=2,
    )

    fig.update_layout(
        barmode="stack",
 
        yaxis_title=None,
        xaxis_title=None,
        xaxis2_title=None,
        template="plotly_dark",
        height=350,
        yaxis=dict(
            categoryorder="array",
            categoryarray=hazard_order
        ),
        legend=dict(
            orientation="h",
            yanchor="bottom",
            y=1.05,
            xanchor="right",
            x=1,
            title_text="EAP?"
        ),
    
        yaxis2=dict(
            showgrid=False,
            showline=True,
            showticklabels=False,

        ),
        
        
        
        margin=dict(l=20, r=20, t=50, b=20),
        
        showlegend=True # Set to False if you want to hide the legend
    )
    

    return fig


def create_tab(df,view, hazard_class):
    
    #FUNCTION CREATES THE DRILLDOWN BY HAZARD POTENTIAL CLASSIFICATION INTO
    #OWNER TYPE OR PRIMARY PURPOSE 
    
    #INPUT DFF2 = OWNER OR DFF6=PURPOSE, ALWAYS FILTERED BY HAZARD POTENTIAL CLASSIFICATION
    
    if view == 'Owner':
        colname = 'Primary Owner Type'
        colnameperc = 'Percentage Hazard Owner'
    else:
        colname = 'Primary Purpose'
        colnameperc = 'Percentage Hazard Purpose'
    
    # Aggregate for the total numbers per view
    dfg = df.groupby(colname)['Number'].sum().reset_index().sort_values(colname, ascending=False)
    
    
    #color subplot if dict will not work and you're in a  hurry and
    #not even create a switch
    
   
    
    if hazard_class == 'High':
        single_bar_color = colors_class[3]
    elif hazard_class == 'Significant':
        single_bar_color = colors_class[2]
    elif hazard_class == 'Low':
        single_bar_color = colors_class[1]
    else:
        single_bar_color = colors_class[0]

    # Desired order and colors
    eap_statuses = ['No', 'Not Required', 'Yes']
    colors = ['rgba(0, 61, 84,1)', 
              'rgba(56, 111, 152, 1)', 
              'rgba(188, 212, 230, 1)']

    # # Set correct order for y-axis
    y_order = df[colname].unique().tolist()[::-1]
    
    df[colname] = pd.Categorical(
        df[colname],
        categories=y_order,
        ordered=True
    )
    df = df.sort_values(colname)
    
        
    # Creating two subplots
    fig = make_subplots(rows=1, cols=2, specs=[[{}, {}]], shared_xaxes=True,
                    shared_yaxes=False, vertical_spacing=0.001,
                    subplot_titles=("Number of dams", "Emergency Action Plan prepared?"))
    
    
    

    # Build traces => subplot 2 % yes, not required, no

    for status, color in zip(eap_statuses, colors):
        subset = df[df["EAP Prepared"] == status]

        fig.add_trace(
            go.Bar(
                y=subset[colname],
                x=subset[colnameperc],
                name=status,
                orientation="h",
                offsetgroup=1,
                marker_color=color,
                text=subset[colnameperc].astype(str) + '%',
                textposition="inside",
                insidetextanchor="middle",

            ),
            row=1,
            col=2,
        ),
     
        
        
        
        
        
      # Subplot 1 - Grouped numbers
    fig.add_trace(
          go.Bar(
             # name=df[colname],
              x=dfg['Number'],
              y=dfg[colname],
              orientation="h",
              showlegend=False,
              marker_color=single_bar_color,
              offsetgroup=0,
              text=dfg['Number'],
                textposition="outside",
                insidetextanchor="end",
              ),
          row=1,
          col=1,
      )      

    

    fig.update_layout(
        barmode="stack",


        
        yaxis_title=None,
        xaxis_title=None,
        xaxis2_title=None,
        template="plotly_dark",
 
        yaxis2=dict(
            showgrid=False,
            showline=True,
            showticklabels=False,

        ),
        legend=dict(
            orientation="h",
            yanchor="bottom",
            y=1.05,
            xanchor="right",
            x=1,
            title_text="EAP?"
        ),
        
        
        
        showlegend=True  # Set to False if you want to hide the legend
    )
    

    return dcc.Graph(figure=fig)











# Initialize Dash app
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP,  dbc_css, vizro_bootstrap])

# Layout
app.layout = dbc.Container([
    dcc.Graph(figure=create_summary(df), id='general-overview'),
    dbc.Row([
        dbc.Col( html.H2(id='click-data')),
        dbc.Col(radioitems)
        ],style={"marginTop":"2rem","marginBottom":"1rem"}),
    html.Div(id='owner-overview')
    
    ])


@app.callback(
    Output('click-data', 'children'),
    Output('owner-overview','children'),
    Input('general-overview', 'clickData'),
    Input("radioitems-input", "value"))
def display_click_data(clickData, value):
    
    if clickData is None:
        hazard_class='High'
    else:
        hazard_class=clickData["points"][0]["label"]
        
    if value is None:
        df=dff2[dff2['Hazard Potential Classification']==hazard_class]
    elif value == "Owner":
        df=dff2[dff2['Hazard Potential Classification']==hazard_class]
    else:
        df=dff6[dff6['Hazard Potential Classification']==hazard_class]
        
                      
                                                           
    h2_title = f"{hazard_class} Hazard Potential, {value}"
    
    return h2_title, create_tab(df, value, hazard_class)

if __name__ == '__main__':
    app.run(debug=True)