EfficientCVaR efficient_risk is throwing error

[dcgithubaccount]

I am using Python3.10 and on mac OS but it seems the problem happens on DataCamp site as well which is using 3.7.

from pypfopt.efficient_frontier import EfficientCVaR ec = EfficientCVaR(None, df, beta=0.9) ec.efficient_risk(target_cvar=0.02) this piece of code is giving error as shown below. It matrix mult is having issues. I think expected_returns should be transposed.

File ~/anaconda3/envs/MortgageModelling/lib/python3.10/site-packages/pypfopt/objective_functions.py:88, in portfolio_return(w, expected_returns, negative) 75 """ 76 Calculate the (negative) mean return of a portfolio 77 (...) 85 :rtype: float 86 """ 87 sign = -1 if negative else 1 ---> 88 mu = w @ expected_returns 89 return _objective_value(w, sign * mu)

[88d52bdba0366127fffca9dfa93895]

Hi @dcgithubaccount,

Could you provide more information to help us identify the source of the error? A minimal code example that reproduces the error and/or the data you are using would be helpful.

We have automatically tested our code with Python 3.10 on the latest macOS and Python 3.7 on the latest Ubuntu, and it is working well.

Regards,

[dcgithubaccount]

Hiya,

Please find the code. The data can be downloaded from https://pastebin.com/UuGKfBAv

udf = pd.read_csv('stock_returns_2018.csv')
from pypfopt.efficient_frontier import EfficientCVaR

ec = EfficientCVaR(None, udf, beta=0.9)
ec.efficient_risk(target_cvar=0.02)

[88d52bdba0366127fffca9dfa93895]

Hi @dcgithubaccount, there is no problem with your data, however, your code needs the expected returns and you let it be None. You can try the following code:

import pandas as pd
from pypfopt import expected_returns
from pypfopt.efficient_frontier import EfficientCVaR

udf = pd.read_csv("~/Downloads/stock_returns_2018.csv.txt")
udf["date"] = pd.to_datetime(udf["date"])
udf.set_index("date", inplace=True)

mu = expected_returns.mean_historical_return(udf, returns_data=True)
ec = EfficientCVaR(mu, udf, beta=0.9)
ec.efficient_risk(target_cvar=0.02)