Given weights w and the covariance matrix covariance of a random vector $X$, return $\mathrm{Var}(w^{\top}X)$.

$$\mathrm{Var}(w^{\top}X) = w^{\top}\Sigma w$$

The off-diagonal terms are the point: correlated inputs can make a portfolio or an ensemble far riskier than summing individual variances suggests.

Input

w = [0.5, 0.5]
covariance = [[1.0, 0.0], [0.0, 1.0]]

Output

0.5