backtest.py

import yfinance as yf
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# class to contain testing and visualization functions
class backtest:

    data = {'Ticker': [],
        'Amount': [],
        'Date': []}
    portfolioVal = []

    def __init__(self) -> None:
        pass

    # function should step through the algorithm time step by time
    # step and record relevant information, including trades and the
    # value of the portfolio at each step
    def run(self, algo):
        for day in algo.ticker.index:     
            algo.decide("AAPL", day)
            self.portfolioVal.append(algo.getCurrVal(day))

    def graphReturns(self, algorithm): 
        df = pd.DataFrame(algorithm.portfolio_vals)

        plt.subplot(2,2,1)
        plt.plot(df.index, df[0], label='Close Price', color='blue')
        plt.title('Portflio values ')
        plt.xlabel('Date')
        plt.ylabel('Price (USD)')
        plt.legend()
        plt.grid(True)

    def graphRSI(self, algorithm): 
        df = pd.DataFrame(algorithm.RSI_vals)

        plt.subplot(2,2,2)
        plt.plot(df.index, df[0], label='Close Price', color='blue')
        plt.title('RSI Values')
        plt.xlabel('Date')
        plt.ylabel('Price (USD)')
        plt.legend()
        plt.grid(True)     

    def calculateVol(self, algorithm):
        df = pd.DataFrame(algorithm.portfolio_vals)
        returns = df.pct_change()
        vols = (returns.std())

        plt.subplot(2,2,3)
        plt.plot(df.index, returns, label='Close Price', color='blue')
        plt.title('Daily Returns')
        plt.xlabel('Date')
        plt.ylabel('Price (USD)')
        plt.legend()
        plt.grid(True)

    def calculate_tot_returns(self, algorithm):
        total_returns = []
        total = 1

        df = pd.DataFrame(algorithm.portfolio_vals)
        returns = df.pct_change()
        returns = returns.dropna()
        for day_return in returns[0]:
            total = total * (1 + day_return)
            total_returns.append(total)

        df2 = pd.DataFrame(total_returns)
        
        plt.subplot(2,2,4)
        plt.plot(df2.index, df2[0], label='Close Price', color='blue')
        plt.title('Normalized Returns')
        plt.xlabel('Date')
        plt.ylabel('Price (USD)')
        plt.legend()
        plt.grid(True)

    def calculate_market_return(self, algorithm):
        start_val = algorithm.market_ticker.iloc[0]["Close"]
        end_val = algorithm.market_ticker.iloc[-1]["Close"]
        market_return = ((end_val / start_val) - 1)*100
        print(f"Market Return Over Time Period: {np.round(market_return, 2)}%")
        start_val = algorithm.portfolio_vals[0]
        end_val = algorithm.portfolio_vals[-1]
        our_return = ((end_val / start_val) - 1)*100
        print(f"Our Return Over Time Period: {np.round(our_return, 2)}%")
        print(f"Performance compared to Market: {np.round(our_return - market_return, 2)}%")