app.py

import streamlit as st
import pandas as pd
import plotly.express as px

st.set_page_config(layout="wide")
# Load the dataset
data = pd.read_csv('manafeth_es_output.csv')

# Remove commas and convert to numeric values for all columns except 'Date'
data[data.columns.difference(['Date'])] = data[data.columns.difference(['Date'])].replace(',', '', regex=True).astype(float)

# Convert 'Date' column to datetime
data['Date'] = pd.to_datetime(data['Date'])

# Set 'Date' column as the index
data.set_index('Date', inplace=True)
data.drop(data.index[-1], inplace=True)


# Filter out rows beyond 2025 and drop them
data.index = pd.to_datetime(data.index)
#data = data[data.index.year <= 2025]
#data.index = data.index.date

# Set up the Streamlit app
st.set_option('deprecation.showPyplotGlobalUse', False)

# # Set the title of your Streamlit app
# title = 'Dashboard'
# # Set the URL of the logo image
# logo_url = "https://i.imgur.com/x5OCu61.png"
# # Create the HTML markup to display the logo
# logo_html = f'<img src="{logo_url}" alt="Logo" style="vertical-align:middle; margin-right:10px; width:50px;">'
# # Combine the logo HTML with the title text
# title_html = f'{logo_html}<span style="font-size:32px;">{title}</span>'
# # Render the title with logo using markdown
# st.markdown(title_html, unsafe_allow_html=True)

# Set the title of your Streamlit app
title = 'Dashboard'
# Set the URL of the logo image
logo_url = "https://i.imgur.com/x5OCu61.png"
# Set the URL of the additional image
image_url = "https://i.imgur.com/xsTBjxI.png"

# Create the HTML markup to display the logo
logo_html = f'<img src="{logo_url}" alt="Logo" style="vertical-align:middle; margin-right:10px; width:50px;">'
# Create the HTML markup to display the additional image
image_html = f'<img src="{image_url}" alt="Image" style="vertical-align:middle; margin-left:625px; width:100px;">'
# Combine the logo HTML, title text, and additional image HTML
title_html = f'{logo_html}<span style="font-size:32px;">{title}</span>{image_html}'
# Render the title with logo and additional image using markdown
st.markdown(title_html, unsafe_allow_html=True)



linewidths = 1300

tab1, tab2, tab3, corrs = st.tabs(["Dataset & Summary", "Plots", "Max Insights",'Borders Comparison & Correlation'])

with tab1:
    # Display the dataset
    with st.container():
        st.dataframe(data)
    # Show summary statistics
    summary = data.copy(deep=True)
    summary.drop('Month sum',axis=1,inplace=True)
    st.subheader('Summary Statistics')
    st.write(summary.describe())
    
import datetime
status = True
with tab2:
    st.title('All borders')
    # Year range selection using two dropdown boxes
    years = pd.unique(data.index.year)
    min_year = min(years)
    max_year = max(years)
    start_year = st.selectbox("Select Start Year", options=range(min_year, max_year + 1), index=0)
    end_year = st.selectbox("Select End Year", options=range(start_year, max_year + 1), index=len(range(start_year, max_year + 1))-1)

    # Convert start_year and end_year to datetime objects
    start_datetime = datetime.datetime(start_year, 1, 1)
    end_datetime = datetime.datetime(end_year, 12, 31)

    # Filter the data based on the selected year range
    filtered_data = data[(data.index >= start_datetime) & (data.index <= end_datetime)]

    with st.expander('All borders sum', expanded=status):
        fig = px.line(filtered_data, x=filtered_data.index, y=filtered_data.columns[-1], title='All borders ' + str(start_year) + '-' + str(end_year))
        fig.update_layout(width=linewidths)
        
        #fig = px.line(data, y=data.columns[-1], color_discrete_sequence=['blue'])
        #fig.add_vrect(x0="2023-05", x1=data.index[-1], fillcolor="rgba(0, 255, 0, 0.3)", layer="below", line_width=0)
        
        st.plotly_chart(fig)

        # Calculate and display the sum of the selected duration
        sum_value = filtered_data.loc[start_datetime:end_datetime, filtered_data.columns[-1]].sum()
        sum_value = "{:,.0f}".format(sum_value)
        st.markdown(f"<p style='color:white;font-size:35px;'>Sum of selected duration: <span style='color:green'>{sum_value}</span></p>", unsafe_allow_html=True)

    st.title('Per border')
    for column in data.columns[:-1]:
        with st.expander(column, expanded=status):
            fig = px.line(filtered_data, x=filtered_data.index, y=column, title=column)
            fig.update_layout(width=linewidths)
            st.plotly_chart(fig)
            
            # Calculate and display the sum of the selected duration
            sum_value = filtered_data.loc[start_datetime:end_datetime, column].sum()
            #st.write('Sum of selected duration:', "{:,.0f}".format(sum_value))
            sum_value = "{:,.0f}".format(sum_value)
            st.markdown(f"<p style='color:white;font-size:35px;'>Sum of selected duration: <span style='color:green'>{sum_value}</span></p>", unsafe_allow_html=True)
            #st.write(f"<p style='color:red;font-size:20px;'>Sum of selected duration: {"{:,.0f}".format(sum_value)}</p>", unsafe_allow_html=True)
            status = False

with tab3:
    col1, col2 = st.columns(2)
    with st.container():
        st.subheader('Stacked Borders')
        # Group data by year and calculate the sum for each border
        yearly_data = summary.groupby(summary.index.year).sum()    
        # Stacked bar chart showing the difference between borders for each year
        fig = px.bar(yearly_data, x=yearly_data.index, y=yearly_data.columns, barmode='stack',
                     title='Stacked Borders by Year')
        fig.update_layout(width=linewidths)
        st.plotly_chart(fig)

    with st.container():
        max_values = summary.max()
        # Converting max_values to a DataFrame with a single column
        max_values_df = pd.DataFrame(max_values, columns=['Maximum Value'])
        # Sorting the DataFrame in descending order
        max_values_sorted = max_values_df.sort_values(by='Maximum Value', ascending=False)
        # Plotting the maximum values over time in descending order
        st.subheader('Top borders')
        fig = px.bar(max_values_sorted, x=max_values_sorted.index, y='Maximum Value', 
                     labels={'x': 'Columns', 'y': 'Maximum Value'}, title='Borders in descending order')
        fig.update_layout(height=600, width=650*2)  # Adjust the height value as desired
        st.plotly_chart(fig)


with corrs:
    st.subheader('Monthly Entries by Border')
    with st.expander('Toggle',expanded=True):
        # Line chart for monthly entries by border
        fig = px.line(data, x=data.index, y=data.columns[:-1], title='Monthly Entries by Border')
        fig.update_layout(width=linewidths)
        st.plotly_chart(fig)
    st.subheader('Correlation heat map')
    with st.expander('Toggle'):
        correlation = summary.corr()
        # st.subheader('Correlation Heatmap')
        # st.write(correlation)
        fig = px.imshow(correlation, color_continuous_scale='RdBu', labels=dict(x='Columns', y='Columns'),
                        title='Correlation Heatmap')
        fig.update_layout(height=600*2,width=600*2)  # Adjust the height value as desired
        st.plotly_chart(fig)