1 2 SUBROUTINE ASSIGN(MM, M, N, A, K, DMU, U, NC, DMAX) 3C 4C<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><> 5C 6C PURPOSE 7C ------- 8C 9C ASSIGNS EACH CASE TO THE CLUSTER WHOSE CENTER IS A MINIMUM 10C EUCLIDEAN DISTANCE FROM THE CASE. CAN BE USED WITH CLUSTER 11C SUBROUTINE RELOC TO FORM A USER-DEFINED K-MEANS PACKAGE. 12C 13C DESCRIPTION 14C ----------- 15C 16C 1. THE VARIABLES MUST BE SCALED SIMILARLY (CLUSTER SUBROUTINE STAND 17C CAN BE USED TO STANDARDIZE THE VARIABLES). 18C 19C 2. THE EUCLIDEAN DISTANCE BETWEEN EACH CLUSTER MEAN AND A CASE IS 20C DETERMINED, AND THAT CASE IS ASSIGNED TO THE CLUSTER WHOSE 21C DISTANCE IS A MINIMUM. THE CASE THAT IS FURTHEST FROM THE 22C CLUSTER CENTER IS RETURNED ALONG WITH THE FURTHEST DISTANCE. 23C MISSING VALUES ARE DENOTED BY A 99999. AND DO NOT AFFECT THE 24C DISTANCE CALCULATION. 25C 26C INPUT PARAMETERS 27C ---------------- 28C 29C MM INTEGER SCALAR (UNCHANGED ON OUTPUT). 30C THE FIRST DIMENSION OF THE MATRIX A. MUST BE AT LEAST M. 31C 32C M INTEGER SCALAR (UNCHANGED ON OUTPUT). 33C THE NUMBER OF CASES. 34C 35C N INTEGER SCALAR (UNCHANGED ON OUTPUT). 36C THE NUMBER OF VARIABLES. 37C 38C A REAL MATRIX WHOSE FIRST DIMENSION MUST BE MM AND WHOSE SECOND 39C DIMENSION MUST BE AT LEAST N (UNCHANGED ON OUTPUT). 40C THE MATRIX OF DATA VALUES. 41C 42C A(I,J) IS THE VALUE FOR THE J-TH VARIABLE FOR THE I-TH CASE. 43C 44C K INTEGER SCALAR (UNCHANGED ON OUTPUT). 45C THE NUMBER OF CLUSTERS. 46C 47C DMU INTEGER SCALAR (UNCHANGED ON OUTPUT). 48C THE FIRST DIMENSION OF THE MATRIX U. MUST BE AT LEAST N. 49C 50C U REAL MATRIX WHOSE FIRST DIMENSION MUST BE DMU AND WHOSE SECOND 51C DIMENSION MUST BE AT LEAST K (UNCHANGED ON OUTPUT). 52C THE CENTERS OF THE CLUSTERS. 53C 54C OUTPUT PARAMETERS 55C ----------------- 56C 57C NC INTEGER VECTOR DIMENSIONED AT LEAST M+1. 58C NC(I) IS THE CLUSTER FOR CASE I. 59C 60C NC(M+1) IS THE CASE FURTHEST FROM ITS MEAN. 61C 62C DMAX REAL SCALAR. 63C THE DISTANCE CASE NC(M+1) IS FROM ITS MEAN. 64C 65C REFERENCE 66C --------- 67C 68C HARTIGAN, J. A. (1975). CLUSTERING ALGORITHMS, JOHN WILEY & 69C SONS, INC., NEW YORK. PAGES 84-112. 70C 71C<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><> 72C 73 74 75