Source code for economic_curves.correlated_brownian_motion.CorBM

import numpy as np 

[docs]def CorBrownian(mu, E, sampleSize): """Algorithm generates samples of increments from a correlated Brownian motion with a given mean and Variance-Covariance matrix (E). The algorithm uses the fact that if you have n independent brownian motions, the samples given by "mu+ C*Z" are distributed as N(mu,E), where mu is the vector of means and C is the square root of the Variance-Covariance matrix. For calculating the square root of the VarCovar matrix, the Cholesky decomposition is implemented. Arguments: mu: Array with n elements containing the mean of each BM E: n x n numpy array with the Variance-Covariance matrix sampleSize: integer representing the number of samples Returns: sampleSize x n numpy array containing sampled increments for the correlated Brownian motion Note: The algorithm is not optimized for speed and no testing of inputs is implemented. If this would be usefull to you, let the authors know so that they can extend the code. Example: >>> import numpy as np >>> mu = [1,2] >>> VarCovar = np.matrix('1,0.8; 0.8,3') >>> sampleSize = 5 >>> CorBrownian(mu,VarCovar, sampleSize) [ 2.83211068 4.50021193] [ 0.26392619 1.56450446] [-0.25928109 0.97167124] [ 1.52038489 1.76274556]] """ def Cholesky(X): """Cholesky–Banachiewicz algorithm decomposes a Hermitian matrix into a product of a lower triangular matrix and its conjugate transpose. Arguments: X: n x n ndarray representing a Hermitian matrix that the user wants to decompose Returns: n x n ndarray lower triangular matrix such that the matrix product between it and its conjugate transpose returns X More info on: https://en.wikipedia.org/wiki/Cholesky_decomposition#The_Cholesky.E2.80.93Banachiewicz_and_Cholesky.E2.80.93Crout_algorithms""" L = np.zeros_like(X) n = X.shape[0] for i in range(0, n): for j in range(0, i+1): sum = 0 for k in range(0, j): sum = sum+ L[i, k]*L[j, k] if (i==j): L[i, j] = np.sqrt(X[i, i]-sum) else: L[i, j] = 1.0/L[j, j] * (X[i, j]-sum) return L dim = E.shape[0] # Guess the number of Brownian motions (dimension) from the size of the Var-Covar matrix Z = np.random.default_rng().normal(0,1,(sampleSize, dim)) # Generate independent increments of a simpleSize dimensional Brownian motion Y = np.zeros((sampleSize, dim)) # Predefine the final output L = Cholesky(E) # Calculate the square root of the Var-Covar matrix for iSample in range(sampleSize): # For each sample, calculate mu + L*Z Y[iSample] =np.transpose(mu) + L @ np.transpose(Z[iSample]) return Y