Binary Search on a Tape
T. C. Hu, Michelle L. Wachs
Abstract
T. C. Hu, Michelle L. Wachs
Abstract
Given n records stored alphabetically on a tape, any comparison search procedure can be characterized by a binary tree. The complete binary tree (binary search) uses the minimum number of comparisons but not the minimum number of movements. The linear binary tree (sequential search) uses the minimum number of movements but not the minimum number of comparisons. A tape-optimal tree is a tree which minimizes the total cost of comparisons and movements. The tape-optimal tree is a “hybrid” of the linear tree and the complete binary tree, and is characterized for arbitrary n.
OpenAlex reports 9 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Given n records stored alphabetically on a tape, any comparison search procedure can be characterized by a binary tree. The complete binary tree (binary search) uses the minimum number of comparisons but not the minimum number of movements. The linear binary tree (sequential search) uses the minimum number of movements but not the minimum number of comparisons. A tape-optimal tree is a tree which minimizes the total cost of comparisons and movements. The tape-optimal tree is a “hybrid” of the linear tree and the complete binary tree, and is characterized for arbitrary n.
Key concepts: Self-balancing binary search tree, Optimal binary search tree, Random binary tree, Interval tree, Ternary search tree, Binary tree, Binary search tree, K-ary tree