prompts.txt

i want to make a beautiful dashboard for household savings using the api call to get the data and a visual to choose time range and the second file to clclualte teh most interesting points and crunch data before sending key points to open ai and a small caht bot on the side to ask more questions import streamlit as st import requests import pandas as pd import openai  # Geocode an address using OpenStreetMap Nominatim def geocode_address(address):     url = "https://nominatim.openstreetmap.org/search"     params = {"q": address, "format": "json"}     headers = {"User-Agent": "streamlit-app"}     resp = requests.get(url, params=params, headers=headers)     resp.raise_for_status()     results = resp.json()     if not results:         return None, None     lon = float(results[0]["lon"])     lat = float(results[0]["lat"])     return lon, lat  # Call Palmetto BEM API with correct parameters def get_bem_data(palmetto_key, lon, lat):     url = "https://ei.palmetto.com/api/v0/bem/calculate"     payload = {         "location": {"longitude": lon, "latitude": lat},         "parameters": {             "from_datetime": "2025-03-19T00:00:00",             "to_datetime": "2025-04-21T00:00:00",             "interval_format": "long",             "group_by": "hour",             "clip_by": "inner",             "variables": "all_non_zero"         }     }     headers = {         "accept": "application/json",         "content-type": "application/json",         "X-API-Key": palmetto_key     }     resp = requests.post(url, json=payload, headers=headers)     resp.raise_for_status()     return resp.json()  # Analyze BEM output with OpenAI to recommend strategies def analyze_with_openai(openai_key, bem_json):     openai.api_key = openai_key     # Extract monthly consumption series     intervals = bem_json.get("data", {}).get("intervals", [])     if not intervals:         return "No interval data to analyze."     # Format prompt     prompt = (         "You are an energy consultant. Given the following monthly energy consumption data for a residential building, "         "provide list of appliances both present and absent  and  5 concise, practical strategies to reduce energy usage.\n\n"         f"Monthly Consumption:\n{intervals}\n\n"         "Strategies:"       )     # Use new OpenAI v1 interface for chat completions     response = openai.chat.completions.create(         model="gpt-4o",         messages=[{"role": "user", "content": prompt}],         max_tokens=3000     )     return response.choices[0].message.content.strip()  # Streamlit UI def main():     st.title("🏠 Residential Energy Analysis & Recommendations")      palmetto_key = st.text_input("Palmetto API Key", type="password")     openai_key = st.text_input("OpenAI API Key", type="password")     address = st.text_input("Enter your street address:")      if st.button("Analyze Energy Data"):         if not all([palmetto_key, openai_key, address]):             st.error("Please provide your Palmetto key, OpenAI key, and an address.")             return          # Geocode         try:             lon, lat = geocode_address(address)         except Exception as e:             st.error(f"Geocoding error: {e}")             return         if lon is None:             st.error("Could not geocode the provided address.")             return         st.success(f"Location: {address} → Lon: {lon:.5f}, Lat: {lat:.5f}")          # Fetch BEM data         try:             bem = get_bem_data(palmetto_key, lon, lat)         except Exception as e:             st.error(f"BEM API error: {e}")             return          # Display monthly consumption         intervals = bem.get("data", {}).get("intervals", [])         if intervals:             df = pd.DataFrame(intervals)             st.subheader("Monthly Energy Consumption (kWh)")             st.dataframe(df)         else:             st.warning("No monthly data returned. Consider requesting additional variables.")          # Generate AI-driven recommendations         try:             recommendations = analyze_with_openai(openai_key, bem)             st.subheader("Recommended Energy-Saving Strategies")             st.markdown(recommendations)         except Exception as e:             st.error(f"OpenAI analysis error: {e}")  if __name__ == "__main__":     main()import streamlit as st import pandas as pd import matplotlib.pyplot as plt import seaborn as sns import numpy as np import plotly.express as px import plotly.graph_objects as go from plotly.subplots import make_subplots  # Set page configuration st.set_page_config(     page_title="Energy Consumption Dashboard",     page_icon="⚡",     layout="wide" )  # Title and instructions st.title("⚡ Energy Consumption Dashboard") st.markdown("Upload your energy consumption CSV file to visualize and analyze your energy usage patterns.")  # File uploader uploaded_file = st.file_uploader("Choose a CSV file", type="csv")  # Function to process and analyze data def analyze_energy_data(df):     # Basic data cleaning and preparation     # Convert datetime strings to datetime objects     df['from_datetime'] = pd.to_datetime(df['from_datetime'])     df['to_datetime'] = pd.to_datetime(df['to_datetime'])          # Create a date column for easier grouping (just the date part)     df['date'] = df['from_datetime'].dt.date          # Extract hour for time-of-day analysis     df['hour'] = df['from_datetime'].dt.hour          # Categorize variables     all_vars = df['variable'].unique()     electricity_vars = [var for var in all_vars if 'electricity' in var]     fossil_fuel_vars = [var for var in all_vars if 'fossil_fuel' in var]     thermal_vars = [var for var in all_vars if 'thermal' in var]     grid_vars = [var for var in all_vars if 'grid' in var]     cost_vars = [var for var in all_vars if 'costs' in var]     emission_vars = [var for var in all_vars if 'emissions' in var]          return df, electricity_vars, fossil_fuel_vars, thermal_vars, grid_vars, cost_vars, emission_vars  # Function to display dashboard def create_dashboard(df, electricity_vars, fossil_fuel_vars, thermal_vars, grid_vars, cost_vars, emission_vars):     # Data overview     st.subheader("📊 Data Overview")          # Key metrics in columns     col1, col2, col3, col4 = st.columns(4)          with col1:         st.metric("Date Range", f"{df['from_datetime'].min().date()} to {df['to_datetime'].max().date()}")          with col2:         st.metric("Total Records", f"{len(df):,}")          with col3:         st.metric("Variables", f"{df['variable'].nunique()}")          with col4:         # Calculate total electricity consumption         if 'consumption.electricity' in df['variable'].values:             total_elec = df[df['variable'] == 'consumption.electricity']['value'].sum()             st.metric("Total Electricity", f"{total_elec:.2f} kWh")         else:             st.metric("Total Electricity", "N/A")          # Tabs for different analyses     tab1, tab2, tab3, tab4 = st.tabs(["Consumption Overview", "Time Patterns", "Detailed Analysis", "Raw Data"])          #################################     # TAB 1: CONSUMPTION OVERVIEW     #################################     with tab1:         st.subheader("Energy Consumption by Category")                  # Total consumption by category         category_totals = df.groupby('variable')['value'].sum().sort_values(ascending=False)                  # Filter for visualization (top categories)         top_categories = category_totals.head(10)                  # Create bar chart using Plotly         fig = px.bar(             x=top_categories.values,             y=top_categories.index,             orientation='h',             title="Top 10 Energy Consumption Categories",             labels={'x': 'Total Consumption', 'y': 'Category'}         )         st.plotly_chart(fig)                  # Component breakdown         st.subheader("Electricity Consumption by Component")                  # Get specific electricity components         elec_components = [var for var in electricity_vars if len(var.split('.')) > 2]                  if elec_components:             component_data = df[df['variable'].isin(elec_components)].groupby('variable')['value'].sum()                          # Create pie chart             fig = px.pie(                 values=component_data.values,                 names=component_data.index,                 title="Electricity Consumption by Component"             )             st.plotly_chart(fig)         else:             st.info("No detailed electricity component data available")          #################################     # TAB 2: TIME PATTERNS     #################################     with tab2:         st.subheader("Energy Consumption Over Time")                  # Daily consumption patterns         main_categories = ['consumption.electricity', 'consumption.fossil_fuel']         available_categories = [cat for cat in main_categories if cat in df['variable'].values]                  if available_categories:             daily_consumption = df[df['variable'].isin(available_categories)].groupby(['date', 'variable'])['value'].sum().reset_index()                          # Create line chart for daily patterns             fig = px.line(                 daily_consumption,                 x='date',                 y='value',                 color='variable',                 title="Daily Energy Consumption",                 labels={'value': 'Consumption', 'date': 'Date', 'variable': 'Energy Type'}             )             st.plotly_chart(fig)                          # Hourly patterns             st.subheader("Hourly Energy Patterns")                          hourly_data = df[df['variable'].isin(available_categories)].groupby(['hour', 'variable'])['value'].mean().reset_index()                          # Create line chart for hourly patterns             fig = px.line(                 hourly_data,                 x='hour',                 y='value',                 color='variable',                 title="Average Hourly Consumption",                 labels={'value': 'Average Consumption', 'hour': 'Hour of Day', 'variable': 'Energy Type'}             )             fig.update_xaxes(tickvals=list(range(0, 24)))             st.plotly_chart(fig)                          # Heatmap             st.subheader("Electricity Consumption Heatmap")                          if 'consumption.electricity' in df['variable'].values:                 elec_hourly = df[df['variable'] == 'consumption.electricity'].copy()                 elec_pivot = elec_hourly.pivot_table(index='date', columns='hour', values='value', aggfunc='mean')                                  # Create heatmap                 fig = px.imshow(                     elec_pivot,                     labels=dict(x="Hour of Day", y="Date", color="Consumption"),                     x=elec_pivot.columns,                     y=elec_pivot.index,                     color_continuous_scale='YlOrRd'                 )                 st.plotly_chart(fig)         else:             st.info("No main category data available for time-based analysis")          #################################     # TAB 3: DETAILED ANALYSIS     #################################     with tab3:         st.subheader("Statistical Summary")                  # Statistical summary for main consumption categories         if available_categories:             stats_cols = st.columns(len(available_categories))                          for i, category in enumerate(available_categories):                 with stats_cols[i]:                     category_data = df[df['variable'] == category]['value']                     st.subheader(f"{category.split('.')[-1].capitalize()}")                     st.metric("Mean", f"{category_data.mean():.3f}")                     st.metric("Median", f"{category_data.median():.3f}")                     st.metric("Min", f"{category_data.min():.3f}")                     st.metric("Max", f"{category_data.max():.3f}")                     st.metric("Std Dev", f"{category_data.std():.3f}")                  # Correlation analysis         st.subheader("Correlation Analysis")                  if len(available_categories) > 1:             # Prepare data for correlation             pivot_data = df[df['variable'].isin(available_categories)].pivot_table(                 index=['from_datetime'],                  columns='variable',                  values='value'             ).reset_index()                          correlation = pivot_data[available_categories].corr()                          # Create heatmap for correlation             fig = px.imshow(                 correlation,                 text_auto=True,                 labels=dict(x="Category", y="Category", color="Correlation"),                 x=correlation.columns,                 y=correlation.index,                 color_continuous_scale='RdBu_r',                 range_color=[-1, 1]             )             st.plotly_chart(fig)                  # Emissions and costs         st.subheader("Emissions and Costs")                  metrics = ['costs', 'emissions']         available_metrics = [m for m in metrics if m in df['variable'].values]                  if available_metrics:             metrics_data = df[df['variable'].isin(available_metrics)].groupby(['date', 'variable'])['value'].sum().reset_index()                          # Create bar chart             fig = px.bar(                 metrics_data,                 x='date',                 y='value',                 color='variable',                 barmode='group',                 title="Daily Costs vs Emissions",                 labels={'value': 'Amount', 'date': 'Date', 'variable': 'Metric'}             )             st.plotly_chart(fig)          #################################     # TAB 4: RAW DATA     #################################     with tab4:         st.subheader("Raw Data")                  # Filter options         st.markdown("### Filter Data")                  # Variable filter         all_vars = sorted(df['variable'].unique())         selected_vars = st.multiselect("Select Variables", all_vars, default=all_vars[:5] if len(all_vars) > 5 else all_vars)                  # Date range filter         date_range = st.date_input(             "Select Date Range",             value=(df['from_datetime'].min().date(), df['to_datetime'].max().date()),             min_value=df['from_datetime'].min().date(),             max_value=df['to_datetime'].max().date()         )                  # Apply filters         filtered_df = df.copy()                  if selected_vars:             filtered_df = filtered_df[filtered_df['variable'].isin(selected_vars)]                  if len(date_range) == 2:             start_date, end_date = date_range             filtered_df = filtered_df[(filtered_df['date'] >= start_date) & (filtered_df['date'] <= end_date)]                  # Display filtered data         st.dataframe(filtered_df)                  # Download link for filtered data         csv = filtered_df.to_csv(index=False)         st.download_button(             label="Download Filtered Data as CSV",             data=csv,             file_name="filtered_energy_data.csv",             mime="text/csv"         )  # Main logic if uploaded_file is not None:     # Load and process the data     df = pd.read_csv(uploaded_file)          # Check if the file has the expected structure     required_columns = ['from_datetime', 'to_datetime', 'variable', 'value']          if all(col in df.columns for col in required_columns):         # Process the data         df, electricity_vars, fossil_fuel_vars, thermal_vars, grid_vars, cost_vars, emission_vars = analyze_energy_data(df)                  # Create the dashboard         create_dashboard(df, electricity_vars, fossil_fuel_vars, thermal_vars, grid_vars, cost_vars, emission_vars)     else:         st.error("The uploaded file doesn't have the required columns. Please ensure it has: from_datetime, to_datetime, variable, and value.") else:     # Sample data message     st.info("👆 Upload a CSV file to begin. Your file should contain columns: from_datetime, to_datetime, variable, and value.")          # Example format     st.markdown("""     ### Expected CSV Format:     ```     ,from_datetime,to_datetime,variable,value     0,2025-04-19T00:00:00,2025-04-19T01:00:00,consumption.electricity.refrigerator,0.0474     1,2025-04-19T01:00:00,2025-04-19T02:00:00,consumption.electricity.refrigerator,0.0462     ...     ```     """)
make it use search to find address ingore premium user and fake data api call veruy important