import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px
from datetime import datetime, timedelta

# Set page config
st.set_page_config(page_title="Healthcare Dashboard", layout="wide")

# Load sample data (replace this with your actual data loading logic)
@st.cache_data
def load_data():
    # Generate sample data
    np.random.seed(42)
    dates = pd.date_range(start="2022-01-01", end="2023-04-30", freq="D")
    departments = ["Cardiology", "Neurology", "Oncology", "Pediatrics", "Emergency"]
    diagnoses = ["Hypertension", "Stroke", "Cancer", "Diabetes", "Pneumonia", "Fracture", "Appendicitis", "Asthma", "Heart Disease", "Influenza"]
    
    data = {
        "Date": np.random.choice(dates, 1000),
        "Patient ID": np.arange(1, 1001),
        "Age": np.random.randint(1, 100, 1000),
        "Gender": np.random.choice(["Male", "Female"], 1000),
        "Department": np.random.choice(departments, 1000),
        "Diagnosis": np.random.choice(diagnoses, 1000),
        "Length of Stay": np.random.randint(1, 30, 1000)
    }
    
    return pd.DataFrame(data)

df = load_data()

# Title and introduction
st.title("Healthcare Dashboard")
st.write("Welcome to the healthcare dashboard. This interactive tool provides insights into patient data, hospital admissions, and key healthcare metrics.")

# Sidebar
st.sidebar.header("Filters")
department = st.sidebar.multiselect("Select Department", options=df["Department"].unique(), default=df["Department"].unique())
date_range = st.sidebar.date_input("Select Date Range", value=(df["Date"].min().date(), df["Date"].max().date()))

# Convert date_range to datetime
start_date = datetime.combine(date_range[0], datetime.min.time())
end_date = datetime.combine(date_range[1], datetime.max.time())

# Filter data based on user selection
filtered_df = df[(df["Department"].isin(department)) & (df["Date"] >= start_date) & (df["Date"] <= end_date)]

# Main content
col1, col2 = st.columns(2)

# Patient Demographics
with col1:
    st.subheader("Patient Demographics")
    fig_gender = px.pie(filtered_df, names="Gender", title="Gender Distribution")
    st.plotly_chart(fig_gender, use_container_width=True)

# Hospital Admissions Over Time
with col2:
    st.subheader("Hospital Admissions Over Time")
    admissions_over_time = filtered_df.groupby("Date").size().reset_index(name="Admissions")
    fig_admissions = px.line(admissions_over_time, x="Date", y="Admissions", title="Daily Hospital Admissions")
    st.plotly_chart(fig_admissions, use_container_width=True)

# Top 10 Diagnoses
st.subheader("Top 10 Diagnoses")
top_diagnoses = filtered_df["Diagnosis"].value_counts().nlargest(10)
fig_diagnoses = px.bar(top_diagnoses, x=top_diagnoses.index, y=top_diagnoses.values, title="Top 10 Diagnoses")
st.plotly_chart(fig_diagnoses, use_container_width=True)

# Average Length of Stay by Department
st.subheader("Average Length of Stay by Department")
avg_los = filtered_df.groupby("Department")["Length of Stay"].mean().sort_values(ascending=True)
fig_los = px.bar(avg_los, x=avg_los.values, y=avg_los.index, orientation="h", title="Average Length of Stay (Days)")
st.plotly_chart(fig_los, use_container_width=True)

# Age Distribution
st.subheader("Age Distribution")
fig_age = px.histogram(filtered_df, x="Age", nbins=20, title="Age Distribution of Patients")
st.plotly_chart(fig_age, use_container_width=True)

# Interactive Data Table
st.subheader("Patient Data")
st.dataframe(filtered_df)

# Add a download button for the filtered data
csv = filtered_df.to_csv(index=False).encode('utf-8')
st.download_button(
    label="Download data as CSV",
    data=csv,
    file_name="healthcare_data.csv",
    mime="text/csv",
)

# Key Metrics
st.subheader("Key Metrics")
col1, col2, col3, col4 = st.columns(4)
col1.metric("Total Patients", len(filtered_df))
col2.metric("Avg. Length of Stay", f"{filtered_df['Length of Stay'].mean():.2f} days")
col3.metric("Most Common Diagnosis", filtered_df["Diagnosis"].mode()[0])
col4.metric("Busiest Department", filtered_df["Department"].value_counts().index[0])