mpr.c

it is regression Algorithm in C/C++.

/*----------------------------------------------------------------------
  File    : mpr.c
  Contents: multivariate polynomial regression
  Author  : Christian Borgelt
  History : 2001.10.13 file created
            2001.10.15 first version completed
            2001.10.18 some messages added, option -v added
            2001.10.20 computation of sum of squared errors added
            2001.10.26 adapted to modified module regress
            2003.08.16 slight changes in error message output
            2003.10.25 logistic regression added (logit transformation)
            2007.02.13 adapted to modified module tabscan
            2007.05.18 execution time measurements added
            2007.10.10 evaluation of attribute directions added
----------------------------------------------------------------------*/
#include <stdio.h>
#include <stdlib.h>
#include <stdarg.h>
#include <string.h>
#include <time.h>
#include <assert.h>
#include <math.h>
#ifndef AS_RDWR
#define AS_RDWR
#endif
#ifndef AS_PARSE
#define AS_PARSE
#endif
#include "io.h"
#ifndef MAT_READ
#define MAT_READ
#endif
#ifndef REG_EXTFN
#define REG_EXTFN
#endif
#include "regress.h"
#ifdef STORAGE
#include "storage.h"
#endif

/*----------------------------------------------------------------------
  Preprocessor Definitions
----------------------------------------------------------------------*/
#define PRGNAME     "mpr"
#define DESCRIPTION "multivariate polynomial regression"
#define VERSION     "version 2.5 (2007.10.10)         " \
                    "(c) 2001-2007   Christian Borgelt"

/* --- error codes --- */
#define OK            0         /* no error */
#define E_NONE        0         /* no error */
#define E_NOMEM     (-1)        /* not enough memory */
#define E_FOPEN     (-2)        /* file open failed */
#define E_FREAD     (-3)        /* file read failed */
#define E_FWRITE    (-4)        /* file write failed */
#define E_OPTION    (-5)        /* unknown option */
#define E_OPTARG    (-6)        /* missing option argument */
#define E_ARGCNT    (-7)        /* wrong number of arguments */
#define E_STDIN     (-8)        /* double assignment of stdin */
#define E_PARSE     (-9)        /* parse errors on input file */
#define E_TARGET   (-10)        /* unknown/invalid target attribute */
#define E_MULTTRG  (-11)        /* multiple target attributes */
#define E_DEGREE   (-12)        /* invalid degree of polynomial */
#define E_DIM      (-13)        /* too few fields */
#define E_LOGIT    (-14)        /* logit transformation failed */
#define E_REGRESS  (-15)        /* regression failed */
#define E_UNKNOWN  (-18)        /* unknown error */

#define SEC_SINCE(t)  ((clock()-(t)) /(double)CLOCKS_PER_SEC)

/*----------------------------------------------------------------------
  Constants
----------------------------------------------------------------------*/
static const char *errmsgs[] = {   /* error messages */
  /* E_NONE      0 */  "no error\n",
  /* E_NOMEM    -1 */  "not enough memory\n",
  /* E_FOPEN    -2 */  "cannot open file %s\n",
  /* E_FREAD    -3 */  "read error on file %s\n",
  /* E_FWRITE   -4 */  "write error on file %s\n",
  /* E_OPTION   -5 */  "unknown option -%c\n",
  /* E_OPTARG   -6 */  "missing option argument\n",
  /* E_ARGCNT   -7 */  "wrong number of arguments\n",
  /* E_STDIN    -8 */  "double assignment of standard input\n",
  /* E_PARSE    -9 */  "parse error(s) on file %s\n",
  /* E_TARGET  -10 */  "unknown or invalid target attribute \"%s\"\n",
  /* E_MULTTRG -11 */  "multiple target attributes\n",
  /* E_DEGREE  -12 */  "invalid degree of polynomial %d\n",
  /* E_DIM     -13 */  "too few fields (need at least 2)\n",
  /* E_LOGIT   -14 */  "logit transformation failed: %g\n",
  /* E_REGRESS -15 */  "regression failed "
                         "(cannot solve equation system)\n",
  /* E_VALUE   -16 */  "file %s, record %d: "
                         "invalid value %s in field %d\n",
  /* E_FLDCNT  -17 */  "file %s, record %d: "
                         "%s%d field(s) instead of %d\n",
  /* E_UNKNOWN -18 */  "unknown error\n"
};

/*----------------------------------------------------------------------
  Global Variables
----------------------------------------------------------------------*/
const  char    *prgname = NULL; /* program name for error messages */
static FILE    *in      = NULL; /* input file */
static FILE    *out     = NULL; /* output file */
static ATTSET  *attset  = NULL; /* attribute set */
static ATTMAP  *attmap  = NULL; /* attribute map */
static TABSCAN *tscan   = NULL; /* table scanner */
static SCAN    *scan    = NULL; /* scanner */
static double  *vec     = NULL; /* vector to read table rows */
static REGRESS *reg     = NULL; /* regression object */

/*----------------------------------------------------------------------
  Functions
----------------------------------------------------------------------*/

static void error (int code, ...)
{                               /* --- print error message */
  va_list    args;              /* list of variable arguments */
  const char *msg;              /* error message */

  assert(prgname);              /* check the program name */
  if (code < E_UNKNOWN) code = E_UNKNOWN;
  if (code < 0) {               /* if to report an error, */
    msg = errmsgs[-code];       /* get the error message */
    if (!msg) msg = errmsgs[-E_UNKNOWN];
    fprintf(stderr, "\n%s: ", prgname);
    va_start(args, code);       /* get variable arguments */
    vfprintf(stderr, msg, args);/* print the error message */
    va_end(args);               /* end argument evaluation */
  }
  #ifndef NDEBUG                /* clean up memory */
  if (vec)    free(vec);        /* and close files */
  if (reg)    reg_deletex(reg, 0);
  if (attmap) am_delete(attmap);
  if (attset) as_delete(attset);
  if (tscan)  ts_delete(tscan);
  if (scan)   sc_delete(scan);
  if (in  && (in  != stdin))  fclose(in);
  if (out && (out != stdout)) fclose(out);
  #endif
  #ifdef STORAGE
  showmem("at end of program"); /* check memory usage */
  #endif
  exit(code);                   /* abort the program */
}  /* error() */

/*--------------------------------------------------------------------*/

int main (int argc, char *argv[])
{                               /* --- main function */
  int     i, k = 0, f;          /* loop variables, counter */
  char    *s;                   /* to traverse options */
  char    **optarg = NULL;      /* option argument */
  char    *fn_dom  = NULL;      /* name of domain file */
  char    *fn_hdr  = NULL;      /* name of table header file */
  char    *fn_tab  = NULL;      /* name of table file */
  char    *fn_reg  = NULL;      /* name of regression function file */
  char    *blanks  = NULL;      /* blanks */
  char    *fldseps = NULL;      /* field  separators */
  char    *recseps = NULL;      /* record separators */
  char    *comment = NULL;      /* comment characters */
  char    *trgname = NULL;      /* target attribute name */
  int     matinp   =  0;        /* flag for numerical matrix input */
  int     flags    = AS_NOXATT|AS_NONULL; /* table file read flags */
  int     dim      = -1;        /* vector dimension */
  int     deg      =  1;        /* degree of polynomial */
  double  logit    = -1;        /* scale for logit transformation */
  int     cnt      =  0;        /* number of data vectors */
  int     verb     =  0;        /* flag for verbose message output */
  int     attcnt   =  0;        /* number of attributes */
  int     tplcnt   =  0;        /* number of tuples */
  double  tplwgt   =  0.0;      /* weight of tuples */
  int     mode     = REG_TITLE; /* description mode */
  TSINFO  *err     = NULL;      /* error information of scanner */
  int     trgid;                /* id of the target column */
  ATT     *att;                 /* to traverse attributes */
  clock_t t;                    /* for time measurements */

  prgname = argv[0];            /* get program name for error msgs. */

  /* --- print startup/usage message --- */
  if (argc > 1) {               /* if arguments are given */
    fprintf(stderr, "%s - %s\n", argv[0], DESCRIPTION);
    fprintf(stderr, VERSION); } /* print a startup message */
  else {                        /* if no argument is given */
    printf("usage: %s [options] [-M|domfile [-d|-h hdrfile]] "
	            "tabfile regfile\n", argv[0]);
    printf("%s\n", DESCRIPTION);
    printf("%s\n", VERSION);
    printf("-M       input is a numerical matrix      "
                    "(default: input is a table)\n");
    printf("-o#      output/target attribute name     "
                    "(default: last attribute)\n");
    printf("-x#      degree of regression polynomial  "
                    "(default: %d, i.e., linear)\n", deg);
    printf("-l#      maximum for logit transformation "
                    "(default: normal regression)\n");
    printf("-e       print exponents of regression variables\n");
    printf("-v       verbose message output "
                    "(print data vector counter)\n");
    printf("-b/f/r#  blank characters, field and record separators\n"
           "         (default: \" \\t\\r\", \" ,\\t\", \"\\n\")\n");
    printf("-C#      comment characters (default: \"#\")\n");
    printf("domfile  file containing domain descriptions\n");
    printf("-d       use default header "
                    "(field names = field numbers)\n");
    printf("-h       read table header (field names) from hdrfile\n");
    printf("hdrfile  file containing table header (field names)\n");
    printf("tabfile  table file to read "
                    "(field names in first record)\n");
    printf("regfile  file to write the regression model to\n");
    return 0;                   /* print a usage message */
  }                             /* and abort the program */

  /* --- evaluate arguments --- */
  for (i = 1; i < argc; i++) {  /* traverse the arguments */
    s = argv[i];                /* get option argument */
    if (optarg) { *optarg = s; optarg = NULL; continue; }
    if ((*s == '-') && *++s) {  /* -- if argument is an option */
      while (1) {               /* traverse characters */