def test_quasi_diag():
prices = get_data()
# compute returns
returns = prices.pct_change().dropna(axis=0, how="all").fillna(0.0)
np.testing.assert_allclose(returns.cov().values,
np.genfromtxt(resource("covariance2.csv")))
np.testing.assert_allclose(returns.corr().values,
np.genfromtxt(resource("correlation2.csv")))
cor = returns.corr().values
links = linkage(dist(cor), method="single")
# uncomment this line if you want to generate a new test resource
# np.savetxt(resource("links.csv"), links, delimiter=",")
np.testing.assert_array_almost_equal(
links, np.loadtxt(resource("links.csv"), delimiter=','))
node = tree(links)
ids = node.pre_order()
assert ids == [
11, 7, 19, 6, 14, 5, 10, 13, 3, 1, 4, 16, 0, 2, 17, 9, 8, 18, 12, 15
]
ordered_tickers = prices.keys()[ids].to_list()
print(ordered_tickers)
assert ordered_tickers == [
'UAA', 'WMT', 'SBUX', 'AMD', 'RRC', 'GE', 'T', 'XOM', 'BABA', 'AAPL',
'AMZN', 'MA', 'GOOG', 'FB', 'PFE', 'GM', 'BAC', 'JPM', 'SHLD', 'BBY'
]
def hrp(prices, node=None, method="single"):
"""
Computes the expected variance and the weights for the hierarchical risk parity portfolio
:param cov: This is the covariance matrix that shall be used
:param node: Optional. This is the rootnode of the graph describing the dendrogram
:return: variance, weights
"""
returns = prices.pct_change().dropna(axis=0, how="all")
cov, cor = returns.cov(), returns.corr()
links = linkage(dist(cor.values), method=method)
node = node or tree(links)
return __hrp(node, cov.values, weights=np.ones(cov.shape[1]))
def marcos(prices, node=None, method=None):
# make sure the prices are a DataFrame
assert isinstance(prices, pd.DataFrame)
# convert into returns
returns = prices.pct_change().dropna(axis=0, how="all")
# compute covariance matrix and correlation matrices (both as DataFrames)
cov, cor = returns.cov(), returns.corr()
# Compute the root node of the tree
method = method or "ward"
node = node or tree(linkage(dist(cor.values), method=method))
# this is an interesting step
ids = node.pre_order()
# apply bisection, root is now a ClusterNode of the graph
root = bisection(ids=ids)
# It's not clear to me why Marcos is going down this route. Rather than sticking with the graph computed above.
return _hrp(node=root, cov=cov)
def test_dist():
a = np.array([[1.0, 0.2 / np.sqrt(2.0)], [0.2 / np.sqrt(2.0), 1.0]])
np.testing.assert_allclose(dist(a),
np.array([6.552017e-01]),
rtol=1e-6,
atol=1e-6)