1Blurb:: 2Build a locally accurate surrogate from data at a single point 3Description:: 4Local approximations use value, gradient, and possibly Hessian data 5from a single point to form a series expansion for approximating data 6in the vicinity of this point. 7 8The currently available local 9approximation is the \c taylor_series selection. 10 11The truth model to be used to generate the value/gradient/Hessian data 12used in the series expansion is identified through the required \c 13actual_model_pointer specification. The use of a model pointer (as 14opposed to an interface pointer) allows additional flexibility in 15defining the approximation. In particular, the derivative 16specification for the truth model may differ from the derivative 17specification for the approximation, and the truth model results 18being approximated may involve a model recursion (e.g., the 19values/gradients from a nested model). 20 21Topics:: 22Examples:: 23Theory:: 24Faq:: 25See_Also:: model-surrogate-global, model-surrogate-hierarchical, model-surrogate-multipoint 26