Code Editor for app.py

import streamlit as st
import datetime
import pandas as pd

# Initialize session state for filters
if "filters" not in st.session_state:
    st.session_state.filters = []

st.title("Time Range Filter Creator")

# Date range selection
date_range = st.date_input(
    "Select date range",
    value=(datetime.date.today(), datetime.date.today()),
    min_value=datetime.date(2000, 1, 1),
    max_value=datetime.date(2050, 12, 31)
)

# Time range selection
col1, col2 = st.columns(2)

with col1:
    start_time = st.time_input("Start time", datetime.time(9, 0))
    
with col2:
    end_time = st.time_input("End time", datetime.time(17, 0))

if st.button("Add Filter"):
    # Create a filter dictionary
    new_filter = {
        "start_date": date_range[0],
        "end_date": date_range[1],
        "start_time": start_time,
        "end_time": end_time
    }
    
    st.session_state.filters.append(new_filter)
    st.success("Filter added successfully!")

# Display current filters
if st.session_state.filters:
    st.subheader("Current Filters")
    for idx, filter in enumerate(st.session_state.filters):
        st.write(f"Filter {idx + 1}:")
        st.write(f"Date Range: {filter['start_date']} to {filter['end_date']}")
        st.write(f"Time Range: {filter['start_time']} to {filter['end_time']}")
        
    if st.button("Clear All Filters"):
        st.session_state.filters = []
        st.success("All filters cleared!")

# Example of how to apply these filters to a timeseries DataFrame
st.subheader("Example Usage")
st.code("""
# Example of how to apply the filters to your timeseries data:
def apply_filters(df, filters):
    mask = pd.Series(False, index=df.index)
    
    for filter in filters:
        date_mask = (df.index.date >= filter['start_date']) & \
                   (df.index.date <= filter['end_date'])
        
        time_mask = (df.index.time >= filter['start_time']) & \
                   (df.index.time <= filter['end_time'])
        
        mask = mask | (date_mask & time_mask)
    
    return df[mask]
""")