Created
April 21, 2021 12:33
-
-
Save montali/f94d274703efc7401fe3df83690c3a52 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import pandas as pd | |
| import numpy as np | |
| from prophet import Prophet | |
| from art import * | |
| import NelderMead | |
| # We first import the historical data from CSV | |
| historical_data = pd.read_csv('all_stocks_5yr.csv') | |
| historical_data = historical_data.dropna() | |
| # Then, we can make the user choose an investment horizon | |
| tprint("lp-money-machine") | |
| investment_horizon_days = input( | |
| "Welcome. Please, insert an investment horizon in days: ") | |
| def create_prophet_dataframe(stock_data): | |
| """Creates a Prophet-compatible dataframe. Needed because Prophet expects data in a given shape | |
| Args: | |
| stock_data (Pandas dataframe): Historical stock data | |
| Returns: | |
| Pandas dataframe: Prophet-compatible dataframe | |
| """ | |
| prophet_df = pd.DataFrame() | |
| prophet_df["ds"] = stock_data["date"] | |
| prophet_df["y"] = stock_data["close"] | |
| return prophet_df | |
| def predict_stock_return(close_prices, days_from_now): | |
| """Predicts the return and risk over the investment horizon | |
| Args: | |
| close_prices (Pandas DataFrame): Time series containing the closing prices | |
| days_from_now (int): Investment horizon | |
| Returns: | |
| tuple: Return forecast, risk forecast | |
| """ | |
| estimator = Prophet() | |
| estimator.fit(close_prices) | |
| future = estimator.make_future_dataframe(periods=days_from_now) | |
| forecast = estimator.predict(future) | |
| risk = abs(np.std(forecast["yhat"].iloc[-investment_horizon_days:])) | |
| return forecast["yhat"].iloc[-1], risk | |
| # We can now create a Pandas dataframe that will contain our results | |
| stocks_analysis = pd.DataFrame(columns=[ | |
| "Name", "PredictionsFromDate", "PredictionsToDate", "OpenPrice", "Risk", "Prediction"]) | |
| symbols = historical_data["Name"].unique() | |
| for symbol in tqdm(symbols): | |
| open_date = historical_data[historical_data.Name == symbol]["date"].iloc[0] | |
| try: | |
| close_date = historical_data[historical_data.Name == | |
| symbol]["date"].iloc[-1] | |
| # We take the closing price of yesterday as buying price | |
| open_price = historical_data[historical_data.Name == | |
| symbol]["close"].iloc[-1] | |
| prediction, risk = predict_stock_return(create_prophet_dataframe( | |
| historical_data[historical_data.Name == symbol]), investment_horizon_days) | |
| stocks_analysis = stocks_analysis.append({ | |
| "Name": symbol, | |
| "PredictionsFromDate": open_date, | |
| "PredictionsToDate": close_date, | |
| "OpenPrice": open_price, | |
| "Risk": risk, | |
| "Prediction": prediction | |
| }, ignore_index=True) | |
| except IndexError: | |
| print( | |
| f"Index {symbol} doesn't have sufficient historical data for the provided horizon, it will be skipped") | |
| # We now add the ror column, containing predicted return over risk | |
| stocks_analysis["ror"] = (stocks_analysis["Prediction"] - | |
| stocks_analysis["OpenPrice"])/stocks_analysis["Risk"] | |
| stocks_analysis.sort_values("ror", ascending=False) | |
| def objective_function(portfolio): | |
| """The objective function to be optimized | |
| Args: | |
| portfolio (np.array): Portfolio | |
| """ | |
| sum = 0 | |
| for i in range(portfolio): | |
| sum += stocks_analysis.iloc[i]["ror"] * portfolio[i] | |
| return sum | |
| if __name__ == "__main__": | |
| nm = NelderMead(6, objective_function, 1, reflection_parameter=4, expansion_parameter=4, | |
| contraction_parameter=0.05, shrinkage_parameter=0.05) | |
| nm.initialize_simplex() | |
| print(nm.fit(0.00001)) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment