app.py

import os
import streamlit as st
import json
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import matplotlib.animation as animation
import time
from PIL import Image
from streamlit_image_comparison import image_comparison
import numpy as np
#import chromadb

from textwrap import dedent
import google.generativeai as genai

api_key = os.environ["OPENAI_API_KEY"]
from openai import OpenAI

oai_client = OpenAI()

import numpy as np
# Assuming chromadb and TruLens are correctly installed and configured
#from chromadb.utils.embedding_functions import

# Google Langchain
from langchain_google_genai import GoogleGenerativeAI

#Crew imports
from crewai import Agent, Task, Crew, Process

# Retrieve API Key from Environment Variable
GOOGLE_AI_STUDIO = os.environ.get('GOOGLE_API_KEY')

# Ensure the API key is available
if not GOOGLE_AI_STUDIO:
    raise ValueError("API key not found. Please set the GOOGLE_AI_STUDIO2 environment variable.")

# Set gemini_llm
gemini_llm = GoogleGenerativeAI(model="gemini-pro", google_api_key=GOOGLE_AI_STUDIO)



# CrewAI ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++


def crewai_process_gemini(research_topic):
    # Define your agents with roles and goals
    GeminiAgent = Agent(
        role='Story Writer',
        goal='To create a story from bullet points.',
        backstory="""You are an expert writer that understands how to make the average extraordinary on paper """,
        verbose=True,
        allow_delegation=False,
        llm = gemini_llm,
        tools=[
                GeminiSearchTools.gemini_search
                   
      ]

    )


    # Create tasks for your agents
    task1 = Task(
        description=f"""From {research_topic} create your story by writing at least one sentence about each bullet point 
        and make sure you have a transitional statement between scenes . BE VERBOSE.""",
        agent=GeminiAgent
    )

    # Instantiate your crew with a sequential process
    crew = Crew(
        agents=[GeminiAgent],
        tasks=[task1],
        verbose=2,
        process=Process.sequential
    )

    # Get your crew to work!
    result = crew.kickoff()
    
    return result




# Tool import
from crewai.tools.gemini_tools import GeminiSearchTools
from crewai.tools.mixtral_tools import MixtralSearchTools
from crewai.tools.zephyr_tools import ZephyrSearchTools





st.set_page_config(layout="wide")


# Animation Code +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

# Set up the duration for your animation
t0=0 # [hrs]
t_end=2 # [hrs]
dt=0.005 # [hrs]
# Create array for time
t=np.arange(t0,t_end+dt,dt)
frame_amount=len(t)


# Subplot 1
fig2=plt.figure(figsize=(16,9),dpi=120,facecolor=(0.8,0.8,0.8))
gs=gridspec.GridSpec(2,2)
ax0=fig2.add_subplot(gs[0,:],facecolor=(0.9,0.9,0.9))
box_object=dict(boxstyle='circle',fc=(0.1,0.9,0.9),ec='r',lw=10)
stopwatch0=ax0.text(1400,0.65,'',size=20,color='g',bbox=box_object)



def update_plot(num):

    # 1st subplot

    stopwatch0.set_text(str(round(t[num],1))+' hrs')

    return stopwatch0,



# HIN Number +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
from SPARQLWrapper import SPARQLWrapper, JSON
from streamlit_agraph import agraph, TripleStore, Node, Edge, Config
import json

# Function to load JSON data
def load_data(filename):
    with open(filename, 'r') as file:
        data = json.load(file)
    return data

# Dictionary for color codes
color_codes = {
    "residential": "#ADD8E6",
    "commercial": "#90EE90",
    "community_facilities": "#FFFF00",
    "school": "#FFFF00",
    "healthcare_facility": "#FFFF00",
    "green_space": "#90EE90",
    "utility_infrastructure": "#90EE90",
    "emergency_services": "#FF0000",
    "cultural_facilities": "#D8BFD8",
    "recreational_facilities": "#D8BFD8",
    "innovation_center": "#90EE90",
    "elderly_care_home": "#FFFF00",
    "childcare_centers": "#FFFF00",
    "places_of_worship": "#D8BFD8",
    "event_spaces": "#D8BFD8",
    "guest_housing": "#FFA500",
    "pet_care_facilities": "#FFA500",
    "public_sanitation_facilities": "#A0A0A0",
    "environmental_monitoring_stations": "#90EE90",
    "disaster_preparedness_center": "#A0A0A0",
    "outdoor_community_spaces": "#90EE90",
    # Add other types with their corresponding colors
}

# Function to draw the grid with optional highlighting

def draw_grid(data, highlight_coords=None):
    fig, ax = plt.subplots(figsize=(12, 12))
    nrows, ncols = data['size']['rows'], data['size']['columns']
    ax.set_xlim(0, ncols)
    ax.set_ylim(0, nrows)
    ax.set_xticks(range(ncols+1))
    ax.set_yticks(range(nrows+1))
    ax.grid(True)

    # Draw roads with a specified grey color
    road_color = "#606060"  # Light grey; change to "#505050" for dark grey
    for road in data.get('roads', []):  # Check for roads in the data
        start, end = road['start'], road['end']
        # Determine if the road is vertical or horizontal based on start and end coordinates
        if start[0] == end[0]:  # Vertical road
            for y in range(min(start[1], end[1]), max(start[1], end[1]) + 1):
                ax.add_patch(plt.Rectangle((start[0], nrows-y-1), 1, 1, color=road['color']))
        else:  # Horizontal road
            for x in range(min(start[0], end[0]), max(start[0], end[0]) + 1):
                ax.add_patch(plt.Rectangle((x, nrows-start[1]-1), 1, 1, color=road['color']))

    # Draw buildings
    for building in data['buildings']:
        coords = building['coords']
        b_type = building['type']
        size = building['size']
        color = color_codes.get(b_type, '#FFFFFF')  # Default color is white if not specified
        
        if highlight_coords and (coords[0], coords[1]) == tuple(highlight_coords):
            highlighted_color = "#FFD700"  # Gold for highlighting
            ax.add_patch(plt.Rectangle((coords[1], nrows-coords[0]-size), size, size, color=highlighted_color, edgecolor='black', linewidth=2))
        else:
            ax.add_patch(plt.Rectangle((coords[1], nrows-coords[0]-size), size, size, color=color, edgecolor='black', linewidth=1))
            ax.text(coords[1]+0.5*size, nrows-coords[0]-0.5*size, b_type, ha='center', va='center', fontsize=8, color='black')

    ax.set_xlabel('Columns')
    ax.set_ylabel('Rows')
    ax.set_title('Village Layout with Color Coding')
    return fig

# Title ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

    
# Tabs +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

# Create the main app with three tabs
tab1, tab2, tab3 = st.tabs(["Introduction","Green Village", "Control Room"])


with tab1:
    
    st.header("RAGE - A day in the Life of Aya City")

    # Creating columns for the layout
    col1, col2 = st.columns([1, 2])
    
    # Displaying the image in the left column
    with col1:
        image = Image.open('./data/intro_image.jpg')
        st.image(image, caption='Green Open Data City Aya')

    # Displaying the text above on the right
    with col2:
        query = '''
        On his first day at Quantum Data Institute in Green Open Data City Aya, Elian marveled at the city’s harmonious blend of technology and nature. 
        Guided to his mentor, Dr. Maya Lior, a pioneer in urban data ecosystems, their discussion quickly centered on Aya’s innovative design. 
        Dr. Lior explained data analytics and green technologies were intricately woven into the city's infrastructure, and how they used
        a Custom GPT called Green Data City to create the design.
    
        To interact with the Green Data City design tool click the button below.
        '''
        st.markdown(query)
    
        # Displaying the audio player below the text
    
        voice_option = st.selectbox(
        'Choose a voice:',
        ['alloy', 'echo', 'fable', 'onyx', 'nova', 'shimmer']
        )
    
        if st.button('Convert to Speech'):
            if query:
                try:
                    response = oai_client.audio.speech.create(
                        model="tts-1",
                        voice=voice_option,
                        input=query,
                    )
                    
                    # Stream or save the response as needed
                    # For demonstration, let's assume we save then provide a link for downloading
                    audio_file_path = "output.mp3"
                    response.stream_to_file(audio_file_path)
                    
                    # Display audio file to download
                    st.audio(audio_file_path, format='audio/mp3')
                    st.success("Conversion successful!")
                    
                    # Displaying the image with the same name as the selected scene
                    image_file_path = f"./data/{selected_scene}.jpg"  # Adjust the directory as needed
                    try:
                        st.image(image_file_path, caption=f"Scene: {selected_scene}")
                        """All images generated by Dall-E"""
                    except Exception as e:
                        st.error(f"An error occurred while displaying the image: {e}")
                        
                except Exception as e:
                    st.error(f"An error occurred: {e}")
            else:
                st.error("Please enter some text to convert.")
        #st.audio('intro_audio.mp3')
        
        
        st.header("Custom GPT Engineering Tools")
        st.link_button("Create JSON Data for a Green Data Village Population 10,000", "https://chat.openai.com/g/g-4bPJUaHS8-create-a-green-data-village")
        
    st.write("Explanation of the Custom GPT")
    st.write("""
    On clicking "Create Data Village" create a Green Data Village following the 4 Steps below.   Output a JSON file similar to the Example by complete the four Steps.
    To generate the provided JSON code, I would instruct a custom GPT to create a detailed description of a hypothetical smart city layout, named "Green Smart Village", with a population of 10,000. This layout should include a grid size of 21x21, a list of buildings and roads, each with specific attributes:
    
    Step 1:  General Instructions:
    Generate a smart city layout for "Green Smart Village" with a 21x21 grid. Include a population of 10,000.
    
    Step 2:  Buildings:
    For each building, specify its coordinates on the grid, type (e.g., residential, commercial, healthcare facility), size (in terms of the grid), color, and equipped sensors (e.g., smart meters, water flow sensors).
    Types of buildings should vary and include residential, commercial, community facilities, school, healthcare facility, green space, utility infrastructure, emergency services, cultural facilities, recreational facilities, innovation center, elderly care home, childcare centers, places of worship, event spaces, guest housing, pet care facilities, public sanitation facilities, environmental monitoring stations, disaster preparedness center, outdoor community spaces, typical road, and typical road crossing.
    
    Step 3: Assign each building unique sensors based on its type, ensuring a mix of technology like smart meters, occupancy sensors, smart lighting systems, and environmental monitoring sensors.
    
    Step 4: Roads:
    Detail the roads' start and end coordinates, color, and sensors installed.
    Ensure roads connect significant areas of the city, providing access to all buildings. Equip roads with sensors for traffic flow, smart streetlights, and pollution monitoring.  MAKE SURE ALL BUILDINGS HAVE ACCESS TO A ROAD.
    
    This test scenario would evaluate the model's ability to creatively assemble a smart city plan with diverse infrastructure and technology implementations, reflecting real-world urban planning challenges and the integration of smart technologies for sustainable and efficient city management.
    
    Example: 
    {
      "city": "City Name",
      "population": "Population Size",
      "size": {
        "rows": "Number of Rows",
        "columns": "Number of Columns"
      },
      "buildings": [
        {
          "coords": ["X", "Y"],
          "type": "Building Type",
          "size": "Building Size",
          "color": "Building Color",
          "sensors": ["Sensor Types"]
        }
      ],
      "roads": [
        {
          "start": ["X Start", "Y Start"],
          "end": ["X End", "Y End"],
          "color": "Road Color",
          "sensors": ["Sensor Types"]
        }
      ]
    }
        
        
    
    """)

with tab2:

    st.header("Green Smart Village Application")

    # Divide the page into three columns
    col1, col2, col3 = st.columns(3)
    
    with col1:
        st.header("Today's Agenda")
        st.write("1. Morning Meeting\n2. Review Project Plans\n3. Lunch Break\n4. Site Visit\n5. Evening Wrap-up")
        
        st.header("Agent Advisors")
        st.write("Would you like to optimize your HIN number?")

        # Selection box for the function to execute
        process_selection = st.selectbox(
            'Choose the process to run:',
            ('crewai_process_gemini', 'crewai_process_mixtral_crazy', 'crewai_process_mixtral_normal', 'crewai_process_zephyr_normal', 'crewai_process_phi2')
        )
    
        # Button to execute the chosen function
        if st.button('Run Process'):
            if research_topic:  # Ensure there's a topic provided
                if process_selection == 'crewai_process_gemini':
                    result = crewai_process_gemini(research_topic)
                elif process_selection == 'crewai_process_mixtral_crazy':
                    result = crewai_process_mixtral_crazy(research_topic)
                elif process_selection == 'crewai_process_mixtral_normal':
                    result = crewai_process_mixtral_normal(research_topic)
                elif process_selection == 'crewai_process_zephyr_normal':
                    result = crewai_process_zephyr_normal(research_topic)
                elif process_selection == 'crewai_process_phi2':
                    result = crewai_process_phi2(research_topic)
                st.write(result)
            else:
                st.warning('Please enter a research topic.')
            
            st.header("My Incentive")
            st.write("Total incentive for HIN optimization")
    
    with col2:
        st.header("Green Smart Village Layout")
        data = load_data('grid.json')  # Ensure this path is correct
        
        # Dropdown for selecting a building
        building_options = [f"{bld['type']} at ({bld['coords'][0]}, {bld['coords'][1]})" for bld in data['buildings']]
        selected_building = st.selectbox("Select a building to highlight:", options=building_options)
        selected_index = building_options.index(selected_building)
        selected_building_coords = data['buildings'][selected_index]['coords']

        # Draw the grid with the selected building highlighted
        fig = draw_grid(data, highlight_coords=selected_building_coords)
        st.pyplot(fig)
        
        # Assuming sensors are defined in your data, display them
        sensors = data['buildings'][selected_index].get('sensors', [])
        st.write(f"Sensors in selected building: {', '.join(sensors)}")
        
    with col3:
        st.header("Check Your HIN Number")
        # config = Config(height=400, width=400, nodeHighlightBehavior=True, highlightColor="#F7A7A6", directed=True, collapsible=True)


        if sensors:  # Check if there are sensors to display
            graph_store = TripleStore()
            building_name = f"{data['buildings'][selected_index]['type']} ({selected_building_coords[0]}, {selected_building_coords[1]})"
            
            # Iterate through each sensor and create a triple linking it to the building
            for sensor in sensors:
                sensor_id = f"Sensor: {sensor}"  # Label for sensor nodes
                # Correctly add the triple without named arguments
                graph_store.add_triple(building_name, "has_sensor", sensor_id)
        
            # Configuration for the graph visualization
            agraph_config = Config(height=300, width=300, nodeHighlightBehavior=True, highlightColor="#F7A7A6", directed=True, collapsible=True)
        
            # Display the graph
            agraph(nodes=graph_store.getNodes(), edges=graph_store.getEdges(), config=agraph_config)

    

        hin_number = st.text_input("Enter your HIN number:")
        if hin_number:
            st.write("HIN number details...")  # Placeholder for actual HIN number check

with tab3:
   
    st.header("Control Room")
    st.write("Synthetic data should be used to drive control room")



    """
    Smart meters
    
    Water flow sensors
    
    Temperature and humidity sensors
    
    Occupancy sensors
    
    HVAC control systems
    
    Smart lighting
    
    Security cameras
    
    Indoor air quality sensors
    
    Smart lighting systems
    
    Energy consumption monitors
    
    Patient monitoring systems
    
    Environmental monitoring sensors
    
    Energy management systems
    
    Soil moisture sensors
    
    Smart irrigation systems
    
    Leak detection sensors
    
    Grid monitoring sensors
    
    GPS tracking for vehicles
    
    Smart building sensors
    
    Dispatch management systems
    
    High-speed internet connectivity
    
    Energy consumption monitoring
    
    Smart security systems
    
    Environmental control systems
    
    Security systems
    
    Smart HVAC systems
    
    Smart locks
    
    Water usage monitoring
    
    Smart inventory management systems
    
    Waste level sensors
    
    Fleet management systems for sanitation vehicles
    
    Air quality sensors
    
    Weather stations
    
    Pollution monitors
    
    Early warning systems
    
    Communication networks
    
    Adaptive lighting systems
    
    Traffic flow sensors
    
    Smart streetlights 
    
    """


    """
    Animation Code ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
    
    fig2=plt.figure(figsize=(16,9),dpi=120,facecolor=(0.8,0.8,0.8))
    gs=gridspec.GridSpec(2,2)
    stopwatch_ani=animation.FuncAnimation(fig2,update_plot,frames=frame_amount,interval=20,repeat=True,blit=True)
    st.pyplot(fig2)
    """