OCBA

An python implementation for Optimal computing budget allocation (OCBA) from paper .

It's a function for computing budget allocated to each design with addiction budget given in next round. The first round is not considered in this function, you need to set initional budget for each design and simulate in the calling funcion. When you get the static information after simulation, you can call this function to calculate budget needed for each design in next round. All kinds of situation of statistical information of the designs have been considered, so you can use it with confidence.

How to use

continue_bool, needed_budget = OCBA(design_mean, design_var, design_used, addiction_budget)

For more information,please see the codes.Comment is enough.

Simple test

import numpy as np
from OCBA import OCBA
design_mean = np.arange(1,5)
design_var = np.arange(1,5)
design_used = np.ones(4) * 5
addiction_budget = 10
flag, design_budget = OCBA(design_mean, design_var, design_used, addiction_budget)

output:

flag = True
design_budget = [0. 0. 4. 6.] (numpy.ndarray)