Qiskit 投资组合优化应用程序

Question

I recently got flung into the world of quantum computing and I'm a beginner at coding. I was assigned to do the Portfolio Optimization tutorial of the Qiskit Finance Tutorials and input real data. Truth be told, I'm clueless. It's my understanding that I have to replace the "TICKER" and "RandomDataProvider" parts of the code in order to generate a real-life portfolio.

# Generate expected return and covariance matrix from (random) time-series
stocks = [("TICKER%s" % i) for i in range(num_assets)]
data = RandomDataProvider(tickers=stocks,
                 start=datetime.datetime(2016,1,1),
                 end=datetime.datetime(2016,1,30))
data.run()
mu = data.get_period_return_mean_vector()
sigma = data.get_period_return_covariance_matrix()

I've imported Quandl and WikipediaDataProvider. I want to keep the number of assets the same, using Microsoft "MSFT", Disney "DIS", Nike "NKE", and Home Depot "HD" stocks. How might I apply this financial from Quandl to the tutorial? I've tried this so far:

num_assets = 4

# Generate expected return and covariance matrix from (random) time-series
stocks = [("MSFT%s" , "DIS%s" , "NKE%s" , "HD%s" % i) for i in range(num_assets)]
data = WikipediaDataProvider(tickers=stocks,
                 token="xeesvko2fu6Bt9jg-B1T",
                 start=datetime.datetime(2016,1,1),
                 end=datetime.datetime(2016,1,30))
data.run()
mu = data.get_period_return_mean_vector()
sigma = data.get_period_return_covariance_matrix()

But get the error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-59-19e4d9cde1e3> in <module>
      3 # Generate expected return and covariance matrix from (random) time-series
      4 stocks = [("MSFT%s" , "DIS%s" , "NKE%s" , "HD%s" % i) for i in range(num_assets)]
----> 5 data = WikipediaDataProvider(tickers=stocks,
      6                  token="xeesvko2fu6Bt9jg-B1T",
      7                  start=datetime.datetime(2016,1,1),

TypeError: Can't instantiate abstract class WikipediaDataProvider with abstract methods run

I apologize for my limited coding skills - I'm very new to all of this. Thank you in advance.

Answer 1

The stocks parameter should be a list of strings. If you try:

stocks = ['MSFT', 'DIS', 'NKE', 'HD']

It will work. Just make sure you have the latest Qiskit installed. I ran myself and printed mu and sigma:

mu: [ 0.00057085 -0.00379642  0.00057495 -0.00209479]
sigma: [[0.00059268 0.00036507 0.00022995 0.00025648]
 [0.00036507 0.00041735 0.00016424 0.00027058]
 [0.00022995 0.00016424 0.0002836  0.00022028]
 [0.00025648 0.00027058 0.00022028 0.00042107]]

Answer 2

I signed up for Quandl but it wouldn't give me anything beyond 2016 for free. Yahoo seemed to work for me without any token, and with the latest data.

data = YahooDataProvider(
             tickers = ["AAPL", "MSFT","WORK","TEAM"],
             start=datetime.datetime(2020, 9, 1),
             end=datetime.datetime(2020, 9, 30))