A Martin-Löf test describes a null class whose measures converge to at a computable rate. Dropping the condition on the rate gives a stronger notion: a sequence is weakly 2-random if it avoids every null class. The notion lies strictly between Martin-Löf randomness and 2-randomness, and it has a characterization in terms of computational power: a random sequence is weakly 2-random exactly when it computes no non-computable c.e. set.
2.1 Definition and position
A sequence is weakly 2-random if it belongs to no null class, that is, to no class with uniformly and .
Replacing by , one may assume the classes decreasing, so that tends to .
- Every 2-random sequence is weakly 2-random, and every weakly 2-random sequence is Martin-Löf random.
- No sequence is weakly 2-random. In particular is random but not weakly 2-random.
- There is no universal test for weak 2-randomness: no single null class contains every sequence that is not weakly 2-random.
- A Martin-Löf test describes a null class. Let be null, with decreasing . The relation is , since is left-c.e. uniformly in ; so computes, for each , the least with , which exists since tends to . The classes form a Martin-Löf test relative to containing .
- Let with computable. The classes are uniformly , and their intersection is : a sequence in all of them agrees, for every , with for infinitely many , hence with . This is a null class containing .
- Let be null. Some has measure less than , so its complement is a non-empty class, whose leftmost member is . By (2), it is not weakly 2-random, yet it is outside .
2.2 The Downey–Nies–Weber–Yu theorem
For a Martin-Löf random sequence , the following are equivalent:
- is not weakly 2-random;
- computes a non-computable c.e. set;
- computes a non-computable set.
(2 ⇒ 3) Every c.e. set is .
(3 ⇒ 1) Let be non-computable and , with . The classes are uniformly , and, as in Proposition 2.2 (2), their intersection is , which is null by Sacks’s theorem (Sacks). It contains .
(1 ⇒ 2) Let , a null class with decreasing , and let be the clopen approximation of at stage , generated by the strings enumerated by stage ; it increases with , and . We enumerate a set , together with classes . At stage , for each such that no element has been enumerated for and , if there is with and , enumerate the least such into , and let . If never acts, .
is simple, hence not computable. Each enumerates at most one element, larger than , so has infinite complement. If is infinite, it contains with , since these measures tend to , and then for every . At every stage with , either already meets , or has already acted, or acts; in each case meets (when acts, it enumerates an element of ).
computes . The classes are clopen, uniformly c.e. in , with : a Solovay test. Since is random, belongs to for finitely many only (Solovay), and these enumerate finitely many elements. Let exceed them. For , if enters at stage , then . To decide whether belongs to , search for the least stage with , which exists since ; then if and only if , since increases with .
The construction makes the enumeration of into costly for : it can only happen while is still outside , and the Solovay test charges each such enumeration with the measure of the current approximation. The same idea, charging each enumeration with a measure, underlies the cost functions used to build K-trivial sets.
Item (3) says that, for random sequences, weak 2-randomness amounts to forming a minimal pair with : every set computable both in and in is computable. This formulation is due to Hirschfeldt and Miller.
2.3 Exercises
Show that a weakly 2-random sequence is not computable in , and that a 2-random sequence computes no non-computable c.e. set.
Solution
A set computable in is , hence not weakly 2-random by Proposition 2.2 (2). A 2-random sequence is weakly 2-random and random, so by Theorem 2.3 it computes no non-computable c.e. set.
Show that there is a weakly 2-random sequence that is not 2-random.
Solution
By Corollary 2.5 of “Martin-Löf tests” (3), there is a random sequence that is computably dominated. It computes no non-computable c.e. set : the settling function is computable in , hence in , so there are a computable function and a number with for . Then for , which makes computable. By Theorem 2.3, is weakly 2-random. By Kurtz’s theorem, every 2-random sequence has hyperimmune degree, that is, is not computably dominated; so is not 2-random. Kurtz’s theorem is admitted, as stated in the previous lesson.
2.4 Notes and references
Weak 2-randomness was studied by Kurtz (1981) and by Gaifman and Snir (1982). Downey, Nies, Weber and Yu (2006) proved the characterization of this lesson, and Hirschfeldt and Miller the equivalence with forming a minimal pair with .
- H. Gaifman and M. Snir, Probabilities over rich languages, testing and randomness, Journal of Symbolic Logic 47 (1982), 495–548.
- S. A. Kurtz, Randomness and genericity in the degrees of unsolvability, PhD thesis, University of Illinois, 1981.
- R. Downey, A. Nies, R. Weber and L. Yu, Lowness and Π⁰₂ nullsets, Journal of Symbolic Logic 71 (2006), 1044–1052.
- R. G. Downey and D. R. Hirschfeldt, Algorithmic Randomness and Complexity, Springer, 2010, Section 7.2.