Jackknifing in Xlisp-Stat
Jan de Leeuw · 2011
. It is comparatively easy to write functions in Xlisp-Stat that perform the usual Jackknife computations in any specific problem. It suffices to put a loop around the function we are studying, and to collect the results of the loop in an appropriate way. In this note we use the unique properties of Lisp to automate this process. Contents 1. Introduction 1 2. The Jackknife 2 2.1. Delta Method 2.2. Approximating the Derivatives 3. Some Lisp 3 3.1. Symbols 3.2. Data Types and Function Types 3.3. Evaluation 4. The Jackknife Code 5 4.1. A Pseudovalue Closure 4.2. Pseudovalue Means and Dispersion 4.3. Tying it all together . References 8 1. Introduction In Lisp it is common practice to write functions that return functions. See Tierney [7], Section 3.6.3, and Graham [3], Chapters 5 and 15. Applications of these techniques are, for instance, to produce memoized versions of functions, or versions of function that keep a record of the number of times they are called (which is useful in...