A variant of BWD that reverts to simple Bernoulli randomization if the
imbalance exceeds the threshold.
Public fields
qTarget marginal probability of treatment.
interceptWhether to add an intercept term.
deltaProbability of failure.
NTotal number of points.
DDimension of the data.
value_plusValue added to imbalance when treatment is 1.
value_minusValue added to imbalance when treatment is 0.
phiRobustness parameter.
alphaNormalizing constant (threshold).
w_iCurrent imbalance vector.
iterationsCurrent iteration count.
Active bindings
definitionDictionary of definition parameters.
stateDictionary of current state.
Methods
Method new()
Initialize the BWDRandom balancer.
Usage
BWDRandom$new(N, D, delta = 0.05, q = 0.5, intercept = TRUE, phi = 1)
Arguments
NTotal number of points.
DDimension of the data.
deltaProbability of failure (default 0.05).
qTarget marginal probability of treatment (default 0.5).
interceptWhether to add an intercept term (default TRUE).
phiRobustness parameter (default 1).
Method set_alpha()
Set normalizing constant for remaining N units.
Arguments
NNumber of units remaining.
Method assign_next()
Assign treatment to the next point.
Arguments
xCovariate profile vector.
Returns
Treatment assignment (0 or 1).
Method assign_all()
Assign all points in a matrix (offline setting).
Arguments
XMatrix of covariate profiles (N x D).
Returns
Vector of treatment assignments.
Method update_state()
Update the internal state of the balancer.
Usage
BWDRandom$update_state(w_i, iterations, alpha = NULL)
Arguments
w_iCurrent imbalance vector.
iterationsCurrent iteration count.
alphaOptional. The current threshold value.
Method reset()
Reset the balancer to initial state.
Method clone()
The objects of this class are cloneable with this method.
Usage
BWDRandom$clone(deep = FALSE)
Arguments
deepWhether to make a deep clone.