Commit 84e685a

authored

BUG: Fix RuntimeWarning on dummy compute_stats call (#1253)

Fixes #1251: covariance calculation for insufficient data points

1 parent b1a869c commit 84e685aCopy full SHA for 84e685a

File tree

-2

lines changed

-2

lines changed

Lines changed: 5 additions & 2 deletions

Original file line number	Diff line number	Diff line change
`@@ -154,8 +154,11 @@ def _round_timedelta(value, _period=_data_period(index)):`
`154`	`154`	`s.loc['Calmar Ratio'] = annualized_return / (-max_dd or np.nan)`
`155`	`155`	`equity_log_returns = np.log(equity[1:] / equity[:-1])`
`156`	`156`	`market_log_returns = np.log(c[1:] / c[:-1])`
`157`		`- cov_matrix = np.cov(equity_log_returns, market_log_returns)`
`158`		`- beta = cov_matrix[0, 1] / cov_matrix[1, 1]`
	`157`	`+ beta = np.nan`
	`158`	`+ if len(equity_log_returns) > 1 and len(market_log_returns) > 1:`
	`159`	+ # len == 0 on dummy call `stats_keys = compute_stats(...)` pre optimization
	`160`	`+ cov_matrix = np.cov(equity_log_returns, market_log_returns)`
	`161`	`+ beta = cov_matrix[0, 1] / cov_matrix[1, 1]`
`159`	`162`	`# Jensen CAPM Alpha: can be strongly positive when beta is negative and B&H Return is large`
`160`	`163`	`s.loc['Alpha [%]'] = s.loc['Return [%]'] - risk_free_rate * 100 - beta * (s.loc['Buy & Hold Return [%]'] - risk_free_rate * 100) # noqa: E501`
`161`	`164`	`s.loc['Beta'] = beta`

Comments

(0)