TAOCP/vol2/ch03/01

01

dz / TAOCP / vol2 / ch03 / 01

Summary

3.1: Introduction

Node Tree

Nodes

usefulness_of_random_numbers
content Usefulness of Random Numbers
children recreation, simulation, aesthetics, computer_programming, cryptography, decision_making, numerical_analysis

simulation
content simulation
parents usefulness_of_random_numbers

decision_making
content Decision Making
parents usefulness_of_random_numbers

numerical_analysis
content Numerical Analysis
parents usefulness_of_random_numbers

computer_programming
content Computer Programming
children randomized_algorithms
parents usefulness_of_random_numbers

cryptography
content Cryptography
parents usefulness_of_random_numbers

aesthetics
content Aesthetics
parents usefulness_of_random_numbers

recreation
content Recreation
parents usefulness_of_random_numbers

randomized_algorithms
content Randomized Algorithms
parents computer_programming

seq_rand_num
content Sequence of random independent numbers
children sequence_of_sequences, distribution

distribution
content Distribution
children uniform
parents seq_rand_num

uniform
content Uniform
children equally_probable (description)
parents distribution

equally_probable
content Each number equally probable
parents uniform

sequence_of_sequences
content Sequence of Sequences
parents seq_rand_num

random_number_methods
content Methods to produce random numbers
children computer, mechanical

mechanical
content mechanical
children random_tables
parents random_number_methods

computer
content computer
children random_tables, arithmetic
parents random_number_methods

random_tables
content Random Tables
parents mechanical, computer

arithmetic
content Arithmetic
children neumann_middle_square
parents computer

neumann_middle_square
content Von Neumann's "Middle-square" method
children poor_random_source, deterministic, dont_trust_rng_without_testing (moral of the story)
parents arithmetic

deterministic
content Deterministic
children pseudorandom
parents neumann_middle_square

pseudorandom
content Pseudorandom or Quasi-random
children appear_random
parents deterministic

appear_random
content only appear to be random
parents pseudorandom

poor_random_source
content Comparatively poor source of random numbers
parents neumann_middle_square

dont_trust_rng_without_testing
content middle-square method can give results, but it's dangerous to put much faith into until after elaborate computations have been performed
parents neumann_middle_square
remarks The context here is that many people tried different variations of the middle-square method, and managed to eventually find one that worked really well. But it took lots of testing and "elaborate computations" to get it there.

algo_k
content Alorithm K
children super_random, complicated, converges_6065038420, dont_use_random_methods (lesson learned), knuth_1959, not_very_good

super_random
content "super-random" number generator
parents algo_k

not_very_good
content Not very good at all
parents algo_k

knuth_1959
content Constructed by Knuth in 1959
parents algo_k

complicated
content Complicated
parents algo_k

converges_6065038420
content Coincidentally converges on 6065038420 and transforms back into itself (AKA loops).
parents algo_k

dont_use_random_methods
content Random Number Generators should not be generated with a random method chosen at random
parents algo_k