app.py

import streamlit as st
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import numpy as np
import random

# Function to upload the data from the user
def load_data():
    st.sidebar.header("Upload Your Data")
    uploaded_file = st.sidebar.file_uploader("Choose a CSV or Excel file", type=["csv", "xlsx"])

    if uploaded_file is not None:
        if uploaded_file.name.endswith('.csv'):
            data = pd.read_csv(uploaded_file)
        elif uploaded_file.name.endswith('.xlsx'):
            data = pd.read_excel(uploaded_file)

        # Try to convert date columns to datetime if they exist
        for col in data.columns:
            if 'תכנון' in col or 'ביצוע' in col or 'הגעה' in col:
                try:
                    data[col] = pd.to_datetime(data[col], errors='coerce')  # 'coerce' will convert invalid dates to NaT
                except Exception as e:
                    pass  # skip columns that can't be converted
        return data
    else:
        st.warning("Please upload a CSV or Excel file to proceed.")
        return None

# Function to calculate delay
def calculate_delay(data):
    if 'שעת ביצוע (מבדק)' in data.columns and 'שעת הגעה (מבדק)' in data.columns:
        data['שעת ביצוע (מבדק)'] = pd.to_datetime(data['שעת ביצוע (מבדק)'], errors='coerce')
        data['שעת הגעה (מבדק)'] = pd.to_datetime(data['שעת הגעה (מבדק)'], errors='coerce')
        data['Delay'] = (data['שעת הגעה (מבדק)'] - data['שעת ביצוע (מבדק)']).dt.total_seconds() / 60
    return data

# Function for a generic set of BI questions and plots
def bi_questions_analysis(data):
    st.title("Business Intelligence Dashboard: 60 Questions and Plots")

    # Ensure Delay column is created
    data = calculate_delay(data)

    # Generic function to display random visualizations
    def plot_random_visualization(i):
        random_choice = random.choice(["bar", "line", "scatter", "box", "hist", "heatmap", "pie", "count"])

        # Plot bar chart
        if random_choice == "bar" and 'Delay' in data.columns:
            st.subheader(f"Plot {i}: Bar Chart - Delay by Direction")
            direction_delay = data.groupby('Direction')['Delay'].mean().reset_index()
            fig, ax = plt.subplots()
            sns.barplot(x='Direction', y='Delay', data=direction_delay, ax=ax)
            ax.set_title(f"Delay by Direction")
            st.pyplot(fig)

        # Plot line chart
        elif random_choice == "line" and 'Plantime' in data.columns and 'Delay' in data.columns:
            st.subheader(f"Plot {i}: Line Chart - Delay over Time")
            data['Plantime'] = pd.to_datetime(data['Plantime'], errors='coerce')
            data['hour_of_day'] = data['Plantime'].dt.hour
            daily_delay = data.groupby('hour_of_day')['Delay'].mean()
            fig, ax = plt.subplots()
            daily_delay.plot(kind="line", ax=ax)
            ax.set_title(f"Delay over Time")
            st.pyplot(fig)

        # Plot scatter plot
        elif random_choice == "scatter" and 'Delay' in data.columns:
            st.subheader(f"Plot {i}: Scatter Plot - Delay vs. Plantime")
            fig, ax = plt.subplots()
            sns.scatterplot(x=data['Plantime'], y=data['Delay'], ax=ax)
            ax.set_title(f"Delay vs. Plantime")
            st.pyplot(fig)

        # Plot histogram
        elif random_choice == "hist" and 'Delay' in data.columns:
            st.subheader(f"Plot {i}: Histogram - Distribution of Delay")
            fig, ax = plt.subplots()
            sns.histplot(data['Delay'], kde=True, ax=ax)
            ax.set_title(f"Distribution of Delay")
            st.pyplot(fig)

        # Plot box plot
        elif random_choice == "box" and 'Delay' in data.columns:
            st.subheader(f"Plot {i}: Box Plot - Delay by Status")
            fig, ax = plt.subplots()
            sns.boxplot(x='סטטוס', y='Delay', data=data, ax=ax)
            ax.set_title(f"Delay by Status")
            st.pyplot(fig)

        # Plot pie chart
        elif random_choice == "pie" and 'Delay' in data.columns:
            st.subheader(f"Plot {i}: Pie Chart - Delay Categories")
            delay_status = data['Delay'].apply(lambda x: 'Delayed' if x > 0 else 'On Time').value_counts()
            fig, ax = plt.subplots()
            delay_status.plot(kind="pie", ax=ax, autopct='%1.1f%%', startangle=90)
            ax.set_title(f"Delay Categories")
            st.pyplot(fig)

        # Plot count plot
        elif random_choice == "count" and 'סטטוס' in data.columns:
            st.subheader(f"Plot {i}: Count Plot - Status Frequency")
            fig, ax = plt.subplots()
            sns.countplot(x='סטטוס', data=data, ax=ax)
            ax.set_title(f"Frequency of Status")
            st.pyplot(fig)

        # Plot heatmap
        elif random_choice == "heatmap":
            st.subheader(f"Plot {i}: Heatmap of Correlations")
            corr = data.corr()
            fig, ax = plt.subplots(figsize=(10, 8))
            sns.heatmap(corr, annot=True, fmt=".2f", cmap="coolwarm", ax=ax)
            ax.set_title(f"Correlation Heatmap")
            st.pyplot(fig)

    # Loop through the 12 plots for each page
    for page in range(1, 6):
        st.header(f"Page {page} - 12 Plots")
        for i in range((page - 1) * 12 + 1, page * 12 + 1):
            plot_random_visualization(i)

# Main Page Navigation with buttons for each section
def main_page_navigation(data):
    pages = {
        "Business Intelligence Dashboard": bi_questions_analysis,
    }

    page = st.sidebar.selectbox("Select a page", pages.keys())
    pages[page](data)

# Run the Streamlit app
if __name__ == "__main__":
    data = load_data()

    if data is not None:
        main_page_navigation(data)