macroeconomic_analyser.py

import requests
import yfinance as yf
import pandas as pd
import os
import datetime
from langchain_google_genai import ChatGoogleGenerativeAI
from pytrends.request import TrendReq
from langchain_openai import ChatOpenAI
from langchain_openai import AzureOpenAI
from langchain_groq import ChatGroq
import plotly.graph_objects as go
from plotly.subplots import make_subplots
from dotenv import load_dotenv
import time
import plotly.io as pio
import logging
import warnings
warnings.filterwarnings("ignore", category=FutureWarning)

logging.basicConfig(level=logging.INFO)

pytrends = TrendReq()

load_dotenv(".env")
NEWS_API_API_KEY = os.getenv("NEWS_API_API_KEY")
NEWS_DATA_IO_API_KEY = os.getenv("NEWS_DATA_IO_API_KEY")

countries = {
    'US': '^GSPC',  # S&P 500
    'UK': '^FTSE',  # FTSE 100
    'EU': '^GDAXI',  # DAX
    'China': '000001.SS',  # Shanghai Composite Index
    'India': '^BSESN',  # BSE Sensex
    'Turkey': 'XU100.IS',  # BIST100
    'Japan': '^N225'  # Nikkei 225
}

def _get_indicators(start_date:str, end_date:str) -> tuple:
  
    inflation_rate = get_inflation_rate(start_date=start_date, end_date=end_date)
    
    interest_rates = get_central_bank_interest_rates(start_date=start_date, end_date=end_date)
    
    gdp_growth_rate = get_gdp_growth_rate_last_quarter(start_date=start_date, end_date=end_date)
    trends_data = get_google_trends(['bitcoin', 'ethereum'])
    
    commodities_prices = get_commodities_prices(start_date=start_date, end_date=end_date)
    
    market_indices = get_market_indices(start_date=start_date, end_date=end_date)
    
    currency_strength = get_currency_strength(start_date=start_date, end_date=end_date)


    query_list = [
        'economy AND cryptocurrencies',
        'international conflicts AND war',
        'cryptocurrency regulatory changes OR cryptocurrency news',
        'major cryptocurrency holdings of the corporations OR investments in cryptocurrencies OR ETF approval OR ETF launch',
        'blockchain protocol upgrades OR new blockchain applications OR technological advancements blockchain',
    ]
    queries = []
    for query in query_list:
        articles = _fetch_news(query=query, api_key=NEWS_API_API_KEY, api_key2=NEWS_DATA_IO_API_KEY)
        queries.append(articles)
        time.sleep(1)
    
    macroeconomic_indicator_dataframes = (inflation_rate, interest_rates, gdp_growth_rate, trends_data, commodities_prices, market_indices, currency_strength, queries)
    return macroeconomic_indicator_dataframes


def get_inflation_rate(start_date:str, end_date:str) -> pd.DataFrame:
    # Using yfinance to fetch historical data for consumer price index as a substitute for inflation rates
    cpi_data = {}
    for country, ticker in countries.items():
        data = yf.download(ticker, start=start_date, end=end_date, progress=False)
        if not data.empty:
            # Calculate percentage change in close price as a proxy for inflation rate
            inflation_rate = (data['Close'].iloc[-1] - data['Close'].iloc[0]) / data['Close'].iloc[0] * 100
            cpi_data[country] = inflation_rate

    inflation_rates = pd.DataFrame(list(cpi_data.items()), columns=['Country', 'InflationRate'])
    return inflation_rates


def get_central_bank_interest_rates(start_date:str, end_date:str) -> pd.DataFrame:
    # Using yfinance to fetch market indices data as a substitute for interest rates
    stock_indices = {}
    for country, ticker in countries.items():
        data = yf.download(ticker, start=start_date, end=end_date, progress=False)
        if not data.empty:
            # Calculate average close price as a proxy for interest rates
            avg_close_price = data['Close'].mean()
            stock_indices[country] = avg_close_price

    interest_rates = pd.DataFrame(list(stock_indices.items()), columns=['Country', 'InterestRate'])
    return interest_rates


def get_gdp_growth_rate_last_quarter(start_date:str, end_date:str) -> pd.DataFrame:
    # Using yfinance to fetch historical data for GDP or similar metric to calculate growth rate for last quarter
    gdp_growth_data = {}
    for country, ticker in countries.items():
        data = yf.download(ticker, start=start_date, end=end_date, progress=False)
        if not data.empty:
            # Calculate percentage growth rate for last quarter
            try:
                gdp_growth = (data['Close'].iloc[-1] - data['Close'].iloc[-2]) / data['Close'].iloc[-2] * 100
                gdp_growth_data[country] = gdp_growth
            except IndexError as index_error:
                logging.warning(f"An index error occurred: {index_error}")
            except Exception as e:
                logging.error(f"An error occurred: {e}")
    gdp_growth_rate = pd.DataFrame(list(gdp_growth_data.items()), columns=['Country', 'GDPGrowth'])
    return gdp_growth_rate

def get_google_trends(keywords: list, timeframe:str='now 7-d'):
    pytrends.build_payload(keywords, timeframe=timeframe)
    data = pytrends.interest_over_time()
    return data

def _fetch_news(query:str, api_key:str, api_key2:str)->list:
    page_size = 5
    url = f'https://newsapi.org/v2/everything?q={query}&pageSize={page_size}&apiKey={api_key}'
    response = requests.get(url)
    if response.status_code == 200:
        return response.json()['articles']
    else:
        logging.error(f"Failed to fetch news from NewsAPI: {response.status_code}  {response.text} Trying NewsDataIO...")
        # Fallback to Newsdata.io
        fallback_url = f'https://newsdata.io/api/1/news?apikey={api_key2}&q={query}'
        fallback_response = requests.get(fallback_url)
        if fallback_response.status_code == 200:
            return fallback_response.json()['results']
        else:
            logging.error(f"Failed to fetch news from both newsapi.org and newsdata.io: {fallback_response.status_code}  {fallback_response.text}")
            return []

def get_commodities_prices(start_date:str, end_date:str) -> pd.DataFrame:
    commodities = {
        'Gold': 'GC=F',  # Gold Futures
        'Oil': 'CL=F',   # Crude Oil Futures
        'Silver': 'SI=F', # Silver Futures
        'Copper': 'HG=F', # Copper Futures
        'Wheat': 'ZW=F',  # Wheat Futures
        'Soybeans': 'ZS=F', # Soybean Futures
        'Platinum': 'PL=F', # Platinum Futures
        'Palladium': 'PA=F' # Palladium Futures
    }
    commodities_data = {}
    for commodity, ticker in commodities.items():
        data = yf.download(ticker, start=start_date, end=end_date, progress=False)
        if not data.empty:
            commodities_data[commodity] = data['Close'].pct_change() * 100
    commodities_df = pd.DataFrame(list(commodities_data.items()), columns=['Commodity', 'Price Change'])
    return commodities_df

def get_market_indices(start_date:str,end_date:str) -> pd.DataFrame:
    indices = {
        'S&P 500': '^GSPC',
        'NASDAQ': '^IXIC',
        'DAX': '^GDAXI',
        'FTSE 100': '^FTSE',
        'Nikkei 225': '^N225',
        'Euro Stoxx 50': '^STOXX50E',
        'VIX': '^VIX',        # Volatility Index
        'US 10Y Treasury Yield': '^TNX' # US Treasury Yield 10 Years
    }
    indices_data = {}
    for index, ticker in indices.items():
        data = yf.download(ticker, start=start_date, end=end_date, progress=False)
        if not data.empty:
            indices_data[index] = data['Close'].pct_change() * 100
    indices_df = pd.DataFrame(list(indices_data.items()), columns=['Index', 'Price Change'])
    return indices_df

def get_currency_strength(start_date, end_date) -> pd.DataFrame:
    currency_pairs = {
        'EURUSD': 'EURUSD=X',
        'JPYUSD': 'JPYUSD=X',
        'GBPUSD': 'GBPUSD=X',
        'CNYUSD': 'CNYUSD=X',
        'TRYUSD': 'TRYUSD=X',
        'AUDUSD': 'AUDUSD=X',
        'CHFUSD': 'CHFUSD=X',
        'KRWUSD': 'KRWUSD=X'
    }
    
    exchange_rates = {}
    for pair, ticker in currency_pairs.items():
        data = yf.download(ticker, start=start_date, end=end_date, progress=False)
        if not data.empty:
            exchange_rates[pair[:3]] = data['Close'].pct_change() * 100
    
    currency_df = pd.DataFrame(list(exchange_rates.items()), columns=['Currency', 'Strength Change'])
    return currency_df

def _get_signals(macroeconomic_indicator_dataframes:tuple) -> str:
    inflation_rate, interest_rates, gdp_growth_rate, trends_data, commodities_prices, market_indices, currency_strength, queries = macroeconomic_indicator_dataframes
    
    macroeconomic_signals_and_indicators = "\nMacroeconomic Signals:\n"
    
    macroeconomic_signals_and_indicators += "\nInflation Rate Signals\n"
    for _, row in inflation_rate.iterrows():
        if row['InflationRate'] > 20:
            macroeconomic_signals_and_indicators += f"The inflation rate of {row['Country']} is high at {row['InflationRate']}%. DOWN\n"
        elif row['InflationRate'] < 0:
            macroeconomic_signals_and_indicators += f"The inflation rate of {row['Country']} is low at {row['InflationRate']}%. UP\n"
        else:
            macroeconomic_signals_and_indicators += f"The inflation rate of {row['Country']} is normal at {row['InflationRate']}%. NEUTRAL\n"

    macroeconomic_signals_and_indicators += "\nInterest Rate Signals\n"
    for _, row in interest_rates.iterrows():
        if row['InterestRate'] > 10000:
            macroeconomic_signals_and_indicators += f"The interest rate of {row['Country']} is high at {row['InterestRate']}. UP\n"
        elif row['InterestRate'] < 5000:
            macroeconomic_signals_and_indicators += f"The interest rate of {row['Country']} is low at {row['InterestRate']}. DOWN\n"
        else:
            macroeconomic_signals_and_indicators += f"The interest rate of {row['Country']} is normal at {row['InterestRate']}. NEUTRAL\n"

    macroeconomic_signals_and_indicators += "\nGDP Growth Rate Signals\n"
    for _, row in gdp_growth_rate.iterrows():
        if row['GDPGrowth'] >= 0.5:
            macroeconomic_signals_and_indicators += f"The GDP growth rate of {row['Country']} is high at {row['GDPGrowth']}%. UP\n"
        elif row['GDPGrowth'] < 0.5:
            macroeconomic_signals_and_indicators += f"The GDP growth rate of {row['Country']} is low at {row['GDPGrowth']}%. DOWN\n"
        else:
            macroeconomic_signals_and_indicators += f"The GDP growth rate of {row['Country']} is normal at {row['GDPGrowth']}%. NEUTRAL\n"

    macroeconomic_signals_and_indicators += "\nGoogle Trends Signals\n"
    if trends_data['bitcoin'].iloc[-1] > 100 and trends_data['ethereum'].iloc[-1] > 10:
        macroeconomic_signals_and_indicators += "High interest in Bitcoin and Ethereum. UP\n"
    elif trends_data['bitcoin'].iloc[-1] < 50 and trends_data['ethereum'].iloc[-1] < 5:
        macroeconomic_signals_and_indicators += "Low interest in Bitcoin and Ethereum. DOWN\n"
    else:
        macroeconomic_signals_and_indicators += "Mixed interest in Bitcoin and Ethereum. NEUTRAL\n"

    macroeconomic_signals_and_indicators += "\nMarket Indices Price Signals\n"
    for _, row in market_indices.iterrows():
        price_change = row['Price Change'].iloc[-1]
        if price_change> 10000:
            macroeconomic_signals_and_indicators += f"The price change of {row['Index']} is {price_change}. UP\n"
        elif price_change< 5000:
            macroeconomic_signals_and_indicators += f"The price change of {row['Index']} is {price_change}. DOWN\n"
        else:
            macroeconomic_signals_and_indicators += f"The price change of {row['Index']} is {price_change}. NEUTRAL\n"
    
    macroeconomic_signals_and_indicators += "\nElection Year Signal\n"
    current_year = datetime.datetime.now().year
    if current_year % 4 == 0:
        macroeconomic_signals_and_indicators += f"This year {current_year} is an election year where most of the countries do elections. High Volatility\n"
    else:
        macroeconomic_signals_and_indicators += f"This year {current_year} is not an election year. No comment on Volatility.\n"

    macroeconomic_signals_and_indicators += "\nCommodities Prices Change Signal\n"
    for _, row in commodities_prices.iterrows():
        price_change = row['Price Change'].iloc[-1]
        if price_change >= 0.5:
            macroeconomic_signals_and_indicators += f"The price change of {row['Commodity']} is {price_change}. UP\n"
        elif price_change < 0.5:
            macroeconomic_signals_and_indicators += f"The price change of {row['Commodity']} is {price_change}. DOWN\n"
        else:
            macroeconomic_signals_and_indicators += f"The price change of {row['Commodity']} is {price_change}. NEUTRAL\n"
    
    macroeconomic_signals_and_indicators += "\nCurrency Strength Change Signal\n"
    for _, row in currency_strength.iterrows():
        strength_change = row['Strength Change'].iloc[-1]
        if strength_change > 0.4:
            macroeconomic_signals_and_indicators += f"The strength change of {row['Currency']} is {strength_change}. UP\n"
        elif strength_change < 0.2:
            macroeconomic_signals_and_indicators += f"The strength change of {row['Currency']} is {strength_change}. DOWN\n"
        else:
            macroeconomic_signals_and_indicators += f"The strength change of {row['Currency']} is {strength_change}. NEUTRAL\n"

    macroeconomic_signals_and_indicators += "\nRelated News Articles\n"
    
    for query in queries:
        for news in query:
            macroeconomic_signals_and_indicators += f"{news['title']}.\n"
                
    return macroeconomic_signals_and_indicators

def _get_llm(model_name:str):
    "Return one of the LangChain LLM's according to the model_name. OpenAI, AzureOpenAI, llama-3 and Gemini-pro is supported."
    try:
        if model_name == "openai":
            OPENAI_KEY = os.getenv("OPENAI_KEY")
            llm = ChatOpenAI(model="gpt-4o",temperature=0,max_tokens=None,timeout=None,max_retries=2,api_key=OPENAI_KEY)
        elif model_name == "azure_openai":
            AZURE_AD_TOKEN = os.getenv("AZURE_AD_TOKEN")
            AZURE_AD_TOKEN_PROVIDER = os.getenv("AZURE_AD_TOKEN_PROVIDER")
            AZURE_DEPLOYMENT = os.getenv("AZURE_DEPLOYMENT")
            AZURE_ENDPOINT = os.getenv("AZURE_ENDPOINT")
            llm = AzureOpenAI(azure_ad_token=AZURE_AD_TOKEN, azure_ad_token_provider=AZURE_AD_TOKEN_PROVIDER, azure_deployment=AZURE_DEPLOYMENT, azure_endpoint=AZURE_ENDPOINT, model="gpt-3.5-turbo-instruct", temperature=0,)
        elif model_name == "llama3":
            GROQ_API_KEY = os.getenv("GROQ_API_KEY")
            llm = ChatGroq(api_key=GROQ_API_KEY, model_name="llama3-70b-8192", temperature=0.0) #llama-3.1-70b-versatile
        else:
            GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
            llm = ChatGoogleGenerativeAI(google_api_key=GOOGLE_API_KEY, model="gemini-1.5-pro", temperature=0)
        return llm
    except Exception as e:
        logging.error(f"Error getting LLM: {e}")
        return None

def get_answer_from_llm(llm, signals):
    prompt = """
    You are an experienced financial analyser that focuses on macroeconomical analysis.
    You will be provided a comprehensive analysis of multiple indicators and signals.
    Offer guidance on whether the price go up, down, or remain neutral with the help of these signals.
    Your answer tone should be professional and technical yet easy to understand.
    ONLY ANSWER IN TURKISH.
    
    <Example 1>:
    <HUMAN>:
    Macroeconomic Signals:

    Inflation Rate Signals
    US inflation rate is normal with 19.622963548898117%. NEUTRAL
    UK inflation rate is low with -2.794459435544367%. UP
    EU inflation rate is normal with 10.725078188335043%. NEUTRAL
    China inflation rate is low with -9.806028274777105%. UP
    India inflation rate is high with 20.423220224566734%. DOWN
    Turkey inflation rate is high with 44.78667852482656%. DOWN
    Japan inflation rate is high with 36.20889669289177%. DOWN

    Interest Rate Signals
    US interest rate is low with 4312.705461357694. DOWN
    UK interest rate is normal with 7615.57580953931. NEUTRAL
    EU interest rate is high with 15743.392643229166. UP
    China interest rate is low with 3166.8546579619983. DOWN
    India interest rate is high with 64055.480915178574. UP
    Turkey interest rate is low with 2821.4156112972455. DOWN
    Japan interest rate is high with 30915.395428417174. UP

    GDP Growth Rate Signals
    US GDP growth rate is low with 0.07509755941103574%. DOWN
    UK GDP growth rate is high with 0.637486525155803%. UP
    EU GDP growth rate is high with 0.9520213397443532%. UP
    China GDP growth rate is low with -0.16187801502031762%. DOWN
    India GDP growth rate is high with 1.1813434207185356%. UP
    Turkey GDP growth rate is low with 0.12479279223968566%. DOWN
    Japan GDP growth rate is high with 1.5042856359534253%. UP

    Google Trends Signals
    Mixed interest in Bitcoin and Ethereum. NEUTRAL

    Market Indices Price Signals
    S&P 500 price change is 0.07509755941104146. DOWN
    NASDAQ price change is 0.01716301438894341. DOWN
    DAX price change is 0.9520213397443467. DOWN
    FTSE 100 price change is 0.637486525155806. DOWN
    Nikkei 225 price change is 1.5042856359534174. DOWN
    Euro Stoxx 50 price change is 0.8495690546351353. DOWN
    VIX price change is 2.0900340647264315. DOWN
    US 10Y Treasury Yield price change is -0.6789024496325013. DOWN

    Election Year Signal
    This year 2024 is an election year where most of the countries do elections. High Volatility

    Commodities Prices Change Signal
    Gold price change is 1.6084943433625654. UP
    Oil price change is 0.9164172353136602. UP
    Silver price change is 2.7732172951330902. UP
    Copper price change is -0.9665027734902298. DOWN
    Wheat price change is -1.2836438923395477. DOWN
    Soybeans price change is -2.07106598984772. DOWN
    Platinum price change is 0.12100941539465726. DOWN
    Palladium price change is -0.19500804708108. DOWN

    Currency Strength Change Signal
    EUR strength change is 0.059294316174374195. DOWN
    JPY strength change is 0.4743261914650576. UP
    GBP strength change is 0.2555503062871667. NEUTRAL
    CNY strength change is 0.9474790023711543. UP
    CAD strength change is -0.010470133724260755. DOWN
    AUD strength change is -0.08205419826939897. DOWN
    CHF strength change is -0.059930150097098345. DOWN
    KRW strength change is 0.41996974021509814. UP

    Related News Articles
    Did Altcoin Hunters Miss Meme Coin Season? KAI Cat Says Not Even Close..
    New Play-to-Earn Token Nears $4 Million in Presale, Combining Gaming, Staking, and Crypto.
    Top 3 Crypto Gems with 100x Gains Potential – $PEPU, $CHUANPU, and $PLAY.
    The 7 biggest cryptocurrency myths.
    The Banana Game May Not Be a Scam But It’s Still Not Worth Your Time.
    Israel's many conflicts could soon crack its Iron Dome.
    'Serious breaches' of international law committed in Tigray War, watchdog says.
    Russia's talking about a 6th-generation fighter jet while its Su-57 sits out the Ukraine war.
    Hamas leader believes civilian deaths are ‘necessary sacrifices’ in Israeli war, leaked letters show.
    Meet NATO's next chief, the 'Trump whisperer' and fierce critic of Vladimir Putin.
    Crypto Market Pullback Expected to Bounce Back Soon, PlayDoge Presale Still Going Strong.
    US Presidential debate impacts political memecoin market.
    VanEck SOL ETF Filing – Expert Predicts Solana Price’s Next Move.
    Crypto Resurgence: Navigating The Boom In Cryptocurrency Investments.
    Crypto Firms Must Report Info On Trades To IRS Under New Anti-Tax Evasion Rule.
    Top 3 Meme Coins to Buy in June to 100x Your Investment – $WAI, $DAWGZ, and $PLAY.
    Experts Bullish On Ethereum ETFs Launching July 4, Predicting Potential Surge Beyond $7,000.
    Why FTX’s Generous Bankruptcy Payouts Still Leave Some Creditors Cold.
    US Bitcoin Spot ETFs Milestone: 25,729 BTC Acquired In A Week.
    Bitcoin Firm And University Of Austin Partner For $5 Million Endowment Fund.
    Top Gainer Today: Fetch.ai (FET) Price Keeps Rallying – What Is the Reason for Its Continued Surge?.
    The Backbone of Cybersecurity: Hardware Security Modules.
    Valuing Crypto Assets: Navigating The New Financial Frontier.
    This Altcoin Gem Will Overtake Solana, Predicts Arthur Hayes.
    Solana Eyes $200 Rebound As New Adoption-Driving Technology Debuts.
    
    <AI>:
    Enflasyon Oranı Sinyalleri
    ABD: %19.62 ile normal (NÖTR)
    Birleşik Krallık: % -2.79 ile düşük (YÜKSELİŞ)
    AB: %10.72 ile normal (NÖTR)
    Çin: % -9.81 ile düşük (YÜKSELİŞ)
    Hindistan: %20.42 ile yüksek (DÜŞÜŞ)
    Türkiye: %44.79 ile yüksek (DÜŞÜŞ)
    Japonya: %36.21 ile yüksek (DÜŞÜŞ)
    
    Faiz Oranı Sinyalleri
    ABD: 4312.71 ile düşük (DÜŞÜŞ)
    Birleşik Krallık: 7615.58 ile normal (NÖTR)
    AB: 15743.39 ile yüksek (YÜKSELİŞ)
    Çin: 3166.85 ile düşük (DÜŞÜŞ)
    Hindistan: 64055.48 ile yüksek (YÜKSELİŞ)
    Türkiye: 2821.42 ile düşük (DÜŞÜŞ)
    Japonya: 30915.40 ile yüksek (YÜKSELİŞ)
    
    GSYH Büyüme Oranı Sinyalleri
    ABD: %0.08 ile düşük (DÜŞÜŞ)
    Birleşik Krallık: %0.64 ile yüksek (YÜKSELİŞ)
    AB: %0.95 ile yüksek (YÜKSELİŞ)
    Çin: % -0.16 ile düşük (DÜŞÜŞ)
    Hindistan: %1.18 ile yüksek (YÜKSELİŞ)
    Türkiye: %0.12 ile düşük (DÜŞÜŞ)
    Japonya: %1.50 ile yüksek (YÜKSELİŞ)
    
    Piyasa Endeksleri Fiyat Sinyalleri
    S&P 500: % -0.08 (DÜŞÜŞ)
    NASDAQ: % -0.02 (DÜŞÜŞ)
    DAX: %0.95 (DÜŞÜŞ)
    FTSE 100: %0.64 (DÜŞÜŞ)
    Nikkei 225: %1.50 (DÜŞÜŞ)
    Euro Stoxx 50: %0.85 (DÜŞÜŞ)
    VIX: %2.09 (DÜŞÜŞ)
    ABD 10Y Hazine Tahvili Getirisi: % -0.68 (DÜŞÜŞ)
    Emtia Fiyat Değişim Sinyalleri
    Altın: %1.61 (YÜKSELİŞ)
    Petrol: %0.92 (YÜKSELİŞ)
    Gümüş: %2.77 (YÜKSELİŞ)
    Bakır: % -0.97 (DÜŞÜŞ)
    Buğday: % -1.28 (DÜŞÜŞ)
    Soya Fasulyesi: % -2.07 (DÜŞÜŞ)
    Platin: %0.12 (DÜŞÜŞ)
    Paladyum: % -0.20 (DÜŞÜŞ)
    
    Para Birimi Güç Değişim Sinyalleri
    EUR: % -0.06 (DÜŞÜŞ)
    JPY: %0.47 (YÜKSELİŞ)
    GBP: %0.26 (NÖTR)
    CNY: %0.95 (YÜKSELİŞ)
    TRY: % -0.01 (DÜŞÜŞ)
    AUD: % -0.08 (DÜŞÜŞ)
    CHF: % -0.06 (DÜŞÜŞ)
    KRW: %0.42 (YÜKSELİŞ)
    
    Ek Faktörler
    Google Trends: Bitcoin ve Ethereum’a karmaşık ilgi (NÖTR)
    Seçim Yılı: Birçok ülkede seçimler nedeniyle yüksek volatilite bekleniyor.
    İlgili Haberler: Jeopolitik gerilimler, kripto para gelişmeleri ve ekonomik politika değişikliklerine yüksek odaklanma.
    
    Ekonomik Durum Özeti
    Küresel Enflasyon: ABD ve AB gibi büyük ekonomilerde normal oranlar, ancak Hindistan, Rusya ve Japonya’da yüksek oranlar enflasyonist baskıları gösteriyor.
    Faiz Oranları: AB, Hindistan ve Japonya gibi kilit ekonomilerde yüksek faiz oranları, bu bölgelerde enflasyonu kontrol altına almak için sıkı bir para politikası duruşu olduğunu gösteriyor.
    GSYH Büyümesi: ABD ve Çin hariç çoğu bölgede genel olarak pozitif büyüme, bu iki ülkede ise düşük veya negatif büyüme potansiyel ekonomik yavaşlamalara işaret ediyor.
    Piyasa Endeksleri: Başlıca endekslerde aşağı yönlü eğilimler, ayı piyasası hissiyatı ve potansiyel ekonomik yavaşlamaları gösteriyor.
    Emtialar: Kıymetli metallerde yükseliş eğilimi, güvenli liman varlıklarına yönelimi işaret ederken, bakır gibi endüstriyel emtialarda düşüş eğilimi potansiyel endüstriyel yavaşlamaları gösteriyor.
    Para Birimi Gücü: JPY ve CNY’de belirgin güç, diğer ana para birimlerinde ise zayıflık veya nötr durumlar.
    
    Genel Rehberlik
    Küresel ekonomik durum, bazı bölgelerde ekonomik güç ve büyüme belirtileri görülürken, diğerlerinde yavaşlama ve enflasyonist baskıların olduğu karmaşık sinyallerle karakterizedir.
    Ana endişe alanları arasında birkaç büyük ekonomideki yüksek enflasyon ve piyasa endekslerindeki ayı piyasası eğilimleri yer alıyor. Jeopolitik manzara, yaklaşan seçimler ve jeopolitik gerilimler, küresel ekonomik görünümde ekstra bir volatilite katmanı eklemektedir.
    Yatırım stratejileri, kıymetli metaller gibi güvenli liman varlıklarına odaklanmalı ve sanayi üretimine bağlı hisse senetleri ve emtialarda temkinli olunmalıdır. Bu karmaşık ekonomik ortamda merkez bankası politikalarının ve jeopolitik gelişmelerin izlenmesi kritik olacaktır.
            
    Now it's your turn:
    """
    final_answer = llm.invoke(prompt + signals)
    final_answer = final_answer.content
    return final_answer

def visualize_data(macroeconomic_indicator_dataframes) -> go.Figure:
    inflation_rate, interest_rates, gdp_growth_rate, trends_data, commodities_prices, market_indices, currency_strength, _ = macroeconomic_indicator_dataframes
    
    # Create subplots
    fig = make_subplots(
        rows=3, cols=2,
        subplot_titles=(
            "Inflation Rate", 
            "Interest Rates", 
            "GDP Growth Rate", 
            "Google Trends", 
            "Commodities Prices", 
            "Currency Strength"
        )
    )

    # Inflation Rate
    fig.add_trace(
        go.Scatter(x=inflation_rate['Country'], y=inflation_rate['InflationRate'], mode='lines+markers', name='Inflation Rate'),
        row=1, col=1
    )

    # Interest Rates
    fig.add_trace(
        go.Scatter(x=interest_rates['Country'], y=interest_rates['InterestRate'], mode='lines+markers', name='Interest Rates'),
        row=1, col=2
    )

    # GDP Growth Rate
    fig.add_trace(
        go.Scatter(x=gdp_growth_rate['Country'], y=gdp_growth_rate['GDPGrowth'], mode='lines+markers', name='GDP Growth Rate'),
        row=2, col=1
    )

    # Google Trends
    for keyword in trends_data.columns:
        fig.add_trace(
            go.Scatter(x=trends_data.index, y=trends_data[keyword], mode='lines', name=keyword),
            row=2, col=2
        )

    # Update layout
    fig.update_layout(
        height=900, width=1200, title_text="Macroeconomic Indicators and Signals",
        margin=dict(l=50, r=50, b=50, t=100),  # Adjust margins to improve layout
        showlegend=False  # Hide legend if it causes overlap
    )
    return fig

def macroeconomic_analyser(start_date, end_date, visualization:bool=True) -> tuple[str,str]:
    macroeconomic_indicator_dataframes = _get_indicators(start_date=start_date, end_date=end_date)
    macroeconomic_signals = _get_signals(macroeconomic_indicator_dataframes)
    fig_html = None
    if visualization:
        fig = visualize_data(macroeconomic_indicator_dataframes)
        fig_html = pio.to_html(fig, full_html=True)
    llm = _get_llm(model_name="llama3")
    final_answer = get_answer_from_llm(llm=llm,signals=macroeconomic_signals)
    return final_answer, fig_html

if __name__ == '__main__':
    macroeconomic_analysis = macroeconomic_analyser(start_date='2023-01-13', end_date='2024-01-13')
    print(macroeconomic_analysis)