!************************************************************************* !* * !* Initialization of parameters for LMBM-Clust * !* (version 2.1, last modified 11.11.2017) * !* * !************************************************************************* !* !* Modules included: !* !* initclust ! Initialization of parameters for clustering. !* initlmbm ! Initialization of LMBM -solver. !* MODULE initclust ! Initialization of parameters for clustering codes. USE r_precision, ONLY : prec ! Precision for reals. IMPLICIT NONE ! Names of input and output files: CHARACTER(LEN=80), SAVE :: & infile = 'iris.txt', & ! Name of dataset file. outfile0 = 'centers.txt', & ! Result file with cluster centers. outfile1 = 'indices.txt' ! Result file with function values, ! Davies-Bouldin (DB) and Dunn validity ! indices, and cpu-time. ! Input integer parameters. Give values here. INTEGER, PARAMETER :: & nclust = 25, & ! Maximum number of clusters, from user. nft = 4, & ! Number of features in data, from user. nrecord = 150, & ! Number of recorts in data, from user. maxdim = nclust*nft, & ! Maximum number of variables in optimization. maxsize = 100*maxdim ! Maximum number of candidate points of data set, ! maxsize >> maxdim. ! Input real parameters. Give values here. REAL, SAVE :: & tlimit = 72000.0 ! Maximum CPU-time in seconds, from user. ! Other Real parameters. REAL(KIND=prec), SAVE :: & ! a(nft,nrecord), & ! Data matrix, from input file, give the name of the file above. tnorm, & ! The total number of norms computed. xbest(nclust*nft), & ! Best solution obtained. dcent(nclust,nclust), & ! Distance (affinity) matrix for cluster centers dminim(nrecord) ! ! Other integer parameters. INTEGER, SAVE :: & ! m1, & ! Maximum number of initial solutions. ng1, & ! Number of canditate points ng2, & ! Number of canditate points mf, & ! Number of used features: ! mf = nft, when no classes, nc, & ! Current number of clusters, loops from 1 to nclust ns, & ! Switch for auxiliary and real clustering problem. m, & ! Number of variables in optimization: ! m = mf if ns = 1, ! m = mf*nc if ns = 2. list1(nrecord), & ! list1(i) gives the cluster where point i belongs. nk(nclust,nrecord), & ! nel(nclust), & ! nel(i) = number of records in cluster i. ncand, & ! Number of canditate points. lcand(nrecord) ! CONTAINS SUBROUTINE init_clustpar() ! User supplied subroutine for further initialization of parameters (when needed). ! May be left empty. IMPLICIT NONE END SUBROUTINE init_clustpar SUBROUTINE def_clustpar() ! Default values for parameters. IMPLICIT NONE IF (nrecord <= 1000) THEN m1=500 ng1=8 ng2=9 ELSE IF (nrecord <= 10000) THEN m1=500 ng1=3 ng2=6 ELSE IF (nrecord <= 50000) THEN m1=500 ng1=3 ng2=4 ELSE IF (nrecord <= 100000) THEN m1=300 ng1=1 ng2=2 ELSE m1=200 ng1=1 ng2=2 END IF END SUBROUTINE def_clustpar END MODULE initclust MODULE initlmbm ! Initialization of parameters for LMBM. USE r_precision, ONLY : prec ! Precision for reals. USE param, ONLY : zero, one ! Parameters. USE initclust, ONLY : maxdim ! Maximum dimension of x. IMPLICIT NONE ! Parameters INTEGER, PARAMETER :: & na = 2, & ! Maximum bundle dimension, na >= 2. mcu = 15, & ! Maximum number of stored corrections, mcu >=1. mcinit = 7, & ! Initial maximum number of stored corrections, mcu >= mcinit >= 3. ! If mcinit <= 0, the default value mcinit = 3 will be used. ! However, the value mcinit = 7 is recommented. inma = 3, & ! Selection of line search method: ! inma = 0, Armijo line search, ! inma = 1, nonmonotone Armijo line search. ! inma = 2, weak Wolfe line search. ! inma = 3, nonmonotone weak Wolfe line search. mnma = 10, & ! Maximum number of function values used in nonmonotone line search. maxnin = 20 ! Maximum number of interpolations, maxnin >= 0. ! The value maxnin = 2-20 is recommented with inma=0, ! maxnin >= 20 with inma=1 and 3, and maxnin =200 with inma=2. ! For example: ! inma = 0, maxin = 20. ! inma = 1, mnma = 20, maxin = 30. ! inma = 2, maxnin = 200. ! inma = 3, mnma=10, maxnin = 20. ! Real parameters (if parameter value <= 0.0 the default value of the parameter will be used). REAL(KIND=prec), SAVE :: & tolb = zero, & ! Tolerance for the function value (default = -large). tolf = zero, & ! Tolerance for change of function values (default = 1.0E-8). tolf2 = -10.0_prec, & ! Second tolerance for change of function values. ! - If tolf2 < 0 the the parameter and the corresponding termination ! criterion will be ignored (recommended with inma=1,3). ! - If tolf2 = 0 the default value 1.0E+4 will be used. tolg = 1.0E-5_prec, & ! Tolerance for the termination criterion (default = 1.0E-5). tolg2 = 1.0E-3_prec, & ! Tolerance for the second termination criterion (default = 1.0E-3). eta = 1.0E-4_prec, & ! Distance measure parameter, eta > 0. ! - If eta < 0 the default value 0.0001 will be used. epsl = 0.24E+00, & ! Line search parameter, 0 < epsl < 0.25 (default = 0.24). xmax = 1000.0_prec ! Maximum stepsize, 1 < XMAX (default = 1000). ! Integer parameters (if value <= 0 the default value of the parameter will be used). INTEGER, SAVE :: & n = maxdim, & ! Number of variables. mit = 500, & ! Maximun number of iterations. 500 mfe = 500, & ! Maximun number of function evaluations. 500 mtesf = 0, & ! Maximum number of iterations with changes of ! function values smaller than tolf (default = 10). iprint = 0, & ! Printout specification: ! -1 - No printout. ! 0 - Only the error messages. ! 1 - The final values of the objective function ! (default used if iprint < -1). ! 2 - The final values of the objective function and the ! most serious warning messages. ! 3 - The whole final solution. ! 4 - At each iteration values of the objective function. ! 5 - At each iteration the whole solution iscale = 0 ! Selection of the scaling with LMBM: ! 0 - Scaling at every iteration with STU/UTU (default). ! 1 - Scaling at every iteration with STS/STU. ! 2 - Interval scaling with STU/UTU. ! 3 - Interval scaling with STS/STU. ! 4 - Preliminary scaling with STU/UTU. ! 5 - Preliminary scaling with STS/STU. ! 6 - No scaling. REAL(KIND=prec), DIMENSION(maxdim), SAVE :: x ! Vector of variables CONTAINS SUBROUTINE defaults() ! Default values for parameters. USE param, ONLY: small, large, zero, one, half IMPLICIT NONE IF (iprint < -1) iprint = 1 ! Printout specification. IF (mit <= 0) mit = 500 ! Maximum number of iterations. IF (mfe <= 0) mfe = n*mit ! Maximum number of function evaluations. IF (tolf <= zero) tolf = 1.0E-08_prec ! Tolerance for change of function values. IF (tolf2 == zero) tolf2 = 1.0E+04_prec ! Second tolerance for change of function values. IF (tolb == zero) tolb = -large + small ! Tolerance for the function value. IF (tolg <= zero) tolg = 1.0E-05_prec ! Tolerance for the termination criterion. IF (tolg2 <= zero) tolg = 1.0E-03_prec ! Tolerance for the second termination criterion. IF (xmax <= zero) xmax = 1.5_prec ! Maximum stepsize. IF (eta < zero) eta = half ! Distance measure parameter IF (epsl <= zero) epsl = 1.0E-04_prec ! Line search parameter, IF (mtesf <= 0) mtesf = 10 ! Maximum number of iterations with changes ! of function values smaller than tolf. IF (iscale > 6 .OR. iscale < 0) iscale = 0 ! Selection of the scaling. END SUBROUTINE defaults SUBROUTINE init_lmbmpar() ! User supplied subroutine for further initialization of parameters ! (when needed) for LMBM. May be left empty. USE initclust, ONLY : ns,nc IMPLICIT NONE ! For big data sets use larger values !tolg =1.0E+00 !tolg2 = 1.0E+3 !IF (ns == 2) THEN ! tolg = 1.0E+3 ! IF(nc == 5) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 10) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 15) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 20) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 25) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 30) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 35) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 40) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 45) THEN ! tolg = 1.0E-3 ! ELSE IF(nc == 50) THEN ! tolg = 1.0E-3 ! END IF !END IF ! For small data sets tolg2 = 1.0E-1 IF(ns == 1) tolg = 1.0E-5 IF(ns == 2) tolg = 1.0E-4 END SUBROUTINE init_lmbmpar END MODULE initlmbm