Fractal perturbation preserving fundamental shapes: Bounds on the scale factors

Fractal interpolation function defined through suitable iterated function system provides a method to perturb a function f ∈ C ( I ) so as to yield a class of functions f α ∈ C ( I ) , where α is a free parameter, called scale vector. For suitable values of scale vector α, the fractal functions f α simultaneously interpolate and approximate f. Further, the iterated function system can be selected suitably so that the corresponding fractal function f α shares the quality of smoothness or non-smoothness of f. The objective of the present paper is to choose elements of the iterated function system appropriately in order that f α preserves fundamental shape properties, namely positivity, monotonicity, and convexity in addition to the regularity of f in the given interval. In particular, the scale factors (elements of the scale vector) must be restricted to satisfy two inequalities that provide numerical lower and upper bounds for the multipliers. As a consequence of this process, fractal versions of some elementary theorems in shape preserving interpolation/approximation are obtained. For instance, positive approximation (that is to say, using a positive function) is extended to the fractal case if the factors verify certain inequalities.