app.py

import streamlit as st
import requests
import os
import json
import pandas as pd

# Function to call the Together AI model for the initial analysis
def call_ai_model_initial(all_message):
    url = "https://api.together.xyz/v1/chat/completions"
    payload = {
        "model": "NousResearch/Nous-Hermes-2-Yi-34B",
        "temperature": 1.05,
        "top_p": 0.9,
        "top_k": 50,
        "repetition_penalty": 1,
        "n": 1,
        "messages": [{"role": "user", "content": all_message}],
        "stream_tokens": True,
    }

    TOGETHER_API_KEY = os.getenv('TOGETHER_API_KEY')
    if TOGETHER_API_KEY is None:
        raise ValueError("TOGETHER_API_KEY environment variable not set.")
    
    headers = {
        "accept": "application/json",
        "content-type": "application/json",
        "Authorization": f"Bearer {TOGETHER_API_KEY}",
    }

    response = requests.post(url, json=payload, headers=headers, stream=True)
    response.raise_for_status()  # Ensure HTTP request was successful

    return response

# Function to call the Together AI model for analyzing the text and computing performance score
def call_ai_model_analysis(analysis_text):
    url = "https://api.together.xyz/v1/chat/completions"
    payload = {
        "model": "NousResearch/Nous-Hermes-2-Yi-34B",
        "temperature": 1.05,
        "top_p": 0.9,
        "top_k": 50,
        "repetition_penalty": 1,
        "n": 1,
        "messages": [{"role": "user", "content": analysis_text}],
        "stream_tokens": True,
    }

    TOGETHER_API_KEY = os.getenv('TOGETHER_API_KEY')
    if TOGETHER_API_KEY is None:
        raise ValueError("TOGETHER_API_KEY environment variable not set.")
    
    headers = {
        "accept": "application/json",
        "content-type": "application/json",
        "Authorization": f"Bearer {TOGETHER_API_KEY}",
    }

    response = requests.post(url, json=payload, headers=headers, stream=True)
    response.raise_for_status()  # Ensure HTTP request was successful

    return response

# Streamlit app layout
st.title("Climate Impact on Sports Performance and Infrastructure")
st.write("Analyze and visualize the impact of climate conditions on sports performance and infrastructure.")

# Inputs for climate conditions
temperature = st.number_input("Temperature (°C):", min_value=-50, max_value=50, value=25)
humidity = st.number_input("Humidity (%):", min_value=0, max_value=100, value=50)
wind_speed = st.number_input("Wind Speed (km/h):", min_value=0.0, max_value=200.0, value=15.0)
uv_index = st.number_input("UV Index:", min_value=0, max_value=11, value=5)
air_quality_index = st.number_input("Air Quality Index:", min_value=0, max_value=500, value=100)
precipitation = st.number_input("Precipitation (mm):", min_value=0.0, max_value=500.0, value=10.0)
atmospheric_pressure = st.number_input("Atmospheric Pressure (hPa):", min_value=900, max_value=1100, value=1013)

# Geographic location input
latitude = st.number_input("Latitude:", min_value=-90.0, max_value=90.0, value=0.0)
longitude = st.number_input("Longitude:", min_value=-180.0, max_value=180.0, value=0.0)

if st.button("Generate Prediction"):
    all_message = (
        f"Assess the impact on sports performance and infrastructure based on climate conditions: "
        f"Temperature {temperature}°C, Humidity {humidity}%, Wind Speed {wind_speed} km/h, UV Index {uv_index}, "
        f"Air Quality Index {air_quality_index}, Precipitation {precipitation} mm, Atmospheric Pressure {atmospheric_pressure} hPa. "
        f"Location: Latitude {latitude}, Longitude {longitude}."
        f"After analyzing that, I want you to visualize the data in the best way possible, might be in a table, using a chart or any other way so that it could be easy to understand."
    )

    try:
        with st.spinner("Analyzing climate conditions..."):
            initial_response = call_ai_model_initial(all_message)

            initial_text = ""
            for line in initial_response.iter_lines():
                if line:
                    line_content = line.decode('utf-8')
                    if line_content.startswith("data: "):
                        line_content = line_content[6:]  # Strip "data: " prefix
                    try:
                        json_data = json.loads(line_content)
                        if "choices" in json_data:
                            delta = json_data["choices"][0]["delta"]
                            if "content" in delta:
                                initial_text += delta["content"]
                    except json.JSONDecodeError:
                        continue

            st.success("Initial analysis completed!")

        with st.spinner("Generating predictions..."):
            analysis_text = (
                f"Analyze the following text and extract a performance score based on the climate conditions and their impact: {initial_text}"
            )
            analysis_response = call_ai_model_analysis(analysis_text)

            analysis_result = ""
            for line in analysis_response.iter_lines():
                if line:
                    line_content = line.decode('utf-8')
                    if line_content.startswith("data: "):
                        line_content = line_content[6:]  # Strip "data: " prefix
                    try:
                        json_data = json.loads(line_content)
                        if "choices" in json_data:
                            delta = json_data["choices"][0]["delta"]
                            if "content" in delta:
                                analysis_result += delta["content"]
                    except json.JSONDecodeError:
                        continue

            st.success("Predictions generated!")

            # Extract performance score from the analysis result
            # Assuming the performance score is provided in the text as "Performance Score: XX%"
            performance_score = "N/A"
            for line in analysis_result.split('\n'):
                if "Performance Score:" in line:
                    performance_score = line.split(":")[1].strip()

            # Prepare data for visualization
            results_data = {
                "Condition": ["Temperature", "Humidity", "Wind Speed", "UV Index", "Air Quality Index", "Precipitation", "Atmospheric Pressure"],
                "Value": [temperature, humidity, wind_speed, uv_index, air_quality_index, precipitation, atmospheric_pressure]
            }
            results_df = pd.DataFrame(results_data)

            # Display results in a table
            st.subheader("Results Summary")
            st.table(results_df)

            # Display prediction
            st.markdown("**Predicted Impact on Performance and Infrastructure:**")
            st.markdown(initial_text.strip())

            # Display performance score
            st.markdown(f"**Performance Score:** {performance_score}")

    except ValueError as ve:
        st.error(f"Configuration error: {ve}")
    except requests.exceptions.RequestException as re:
        st.error(f"Request error: {re}")
    except Exception as e:
        st.error(f"An unexpected error occurred: {e}")