1Blurb:: 2(Recommended) Replace points in the surrogate training set, instead of appending 3Description:: 4The user has the option of appending the optimal points from the 5surrogate model to the current set of truth points or using the 6optimal points from the surrogate model to replace the optimal set of 7points from the previous iteration. Although appending to the set is 8the default behavior, at this time we strongly recommend using the 9option \c replace_points because it appears to be more accurate and 10robust. 11 12Topics:: 13Examples:: 14Theory:: 15Faq:: 16See_Also:: 17