**
API
Google Gemini Function Calling**

import yfinance as yf
from vertexai.preview.generative_models import (
FunctionDeclaration,
GenerativeModel,
Part,
Tool,
)

Function to Get Stock Price

def get_stock_price(parameters):
ticker = parameters['ticker']
stock = yf.Ticker(ticker)
hist = stock.history(period="1d")
if not hist.empty:
return {"price": hist['Close'].iloc[-1]}
else:
return {"error": "No data available"}

Tools

tools = Tool(function_declarations=[
FunctionDeclaration(
name="get_stock_price",
description="Get the current stock price of a given company",
parameters={
"type": "object",
"properties": {
"ticker": {
"type": "string",
"description": "Stock ticker symbol"
}
}
},
)
])

Model Initialization

model = GenerativeModel("gemini-pro",
generation_config={"temperature": 0},
tools=[tools])
chat = model.start_chat()

Send a prompt to the chat

prompt = "What is the stock price of Apple?"
response = chat.send_message(prompt)

Check for function call and dispatch accordingly

function_call = response.candidates[0].content.parts[0].function_call

Dispatch table for function handling

function_handlers = {
"get_stock_price": get_stock_price,
}

if function_call.name in function_handlers:
function_name = function_call.name

# Directly extract arguments from function call
args = {key: value for key, value in function_call.args.items()}

# Call the function with the extracted arguments
if args:
    function_response = function_handlers[function_name](args)

    # Sending the function response back to the chat
    response = chat.send_message(
        Part.from_function_response(
            name=function_name,
            response={
                "content": function_response,
            }
        ),
    )

    chat_response = response.candidates[0].content.parts[0].text
    print("Chat Response:", chat_response)
else:
    print("No arguments found for the function.")

else:
print("Chat Response:", response.text)