swh:1:snp:e520bf41b0e99213acde680a9d87fadac1aee079
Tip revision: fab3d29ef16569604858ee648b9e1f6f7d4a7c96 authored by Martin Schlather on 21 September 2014, 00:00:00 UTC
version 3.0.42
version 3.0.42
Tip revision: fab3d29
MLE.cc
/*
Authors
Martin Schlather, schlather@math.uni-mannheim.de
library for simulation of random fields
Copyright (C) 2001 -- 2014 Martin Schlather,
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 3
of the License, or (at your option) any later version.
RO
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*/
#include <R.h>
#include <Rdefines.h>
#include <R_ext/Linpack.h>
#include <math.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "RF.h"
#include "primitive.h"
// #include "Covariance.h"
bool is_top(cov_model *cov) {
if (cov == NULL) BUG;
return isInterface(cov);
}
void GetNAPosition(cov_model *cov,
int *NAs, naptr_type mem, int *elmnts, elptr_type mem_elmnts,
NAname_type names, sortsofparam *sorts, bool *isnan,
covptr_type covModels,
int *covzaehler, int allowforintegerNA,
int SHORTlen, int printing, int depth, bool no_variance,
bool excludetrends
) {
/*
printing <= 0 nothing
1 only NA, NaN
2 others as "*"
3 others as value
mem[0..NAs] : addresse wo (neue) NA position im Speicher
sorts[0..NAs] : see sortsofparam in RF.h
isnan[0..NAs] : NA or NaN
*/
// PMI(cov);
// if (SHORTlen < 2) BUG;
int i, c, r,
namenr = 1,
*nrow = cov->nrow,
*ncol = cov->ncol;
// double *lastmem = NULL;
char shortname[255], shortD[255];
cov_fct *C = CovList + cov->nr, // nicht gatternr
*CC = C;
SEXPTYPE *type = C->kappatype;
if (isRandom(cov)) {
error("Random functions not allowed in models of RFfit, yet");
}
//printf("depth =%d %s %d\n", depth, C->name, *elmnts);
// C is pointer to true model; CC points to the first model passed
// in initNerror.cc by IncludeModel or IncludePrim
while(strncmp(CC->name, InternalName, strlen(InternalName)) ==0) CC--;
covzaehler[cov->nr]++;
//
// print("SHORT %d %s\n", SHORTlen, CC->name);
strcopyN(shortname, CC->nick + 2, SHORTlen);
if (covzaehler[cov->nr] >= 2) {
char dummy[255];
strcopyN(dummy, shortname, SHORTlen-1);
sprintf(shortname, "%s%d", dummy, covzaehler[cov->nr]);
}
if (printing>0) PRINTF("%s\n", CC->name);
// CC needed below for the kappa.names which are given
// print("%d:%s %d %s\n", cov->nr, CC->name, covzaehler[cov->nr], shortname);
depth++;
for (i=0; i<C->kappas; i++) {
if (nrow[i] == 0 || ncol[i] == 0) continue;
if (printing > 0) {
leer(depth); PRINTF("%s\n", C->kappanames[i]);
}
if (i==0 && type[i] == INTSXP && strcmp(C->kappanames[i], ELEMENT) == 0){
if (*elmnts >= MAX_MLE_ELMNTS )
ERR("maximum number of models with variable elements reached");
if (P0INT(i) == NA_INTEGER) ERR1("'%s' may not be NA", ELEMENT);
mem_elmnts[(*elmnts)++] = PINT(i);
// to do!! covModels[*NAs] hierher uebertragen !! s.u.
//printf("elmnts %s %d %s %d\n", CC->name, i, C->kappanames[i], *elmnts);
continue;
}
for (r=0; r<nrow[i]; r++) {
int nv = 0; // anzahl NA in aktuellem parameter
for (c=0; c<ncol[i]; c++) {
double v; // value in aktuellem parameter
int idx = c * nrow[i] + r;
if (*NAs >= MAX_NA) ERR("maximum number of NA reached");
// print("NAs=%d, %d %d %s\n", *NAs, i, idx, C->kappanames[i]);
if (type[i] == REALSXP) {
v = P(i)[idx];
mem[*NAs] = P(i) + idx;
covModels[*NAs] = cov;
} else if (type[i] == INTSXP) {
v = PINT(i)[idx] == NA_INTEGER
? RF_NA : (double) PINT(i)[idx];
if (ISNAN(v) && !allowforintegerNA) {
// crash(cov);
ERR("integer variables currently do not allow for NA"); // !!!
}
// mem[*NAs] = (*double) &(P0INT(i])[idx]);
} else if (type[i] == LISTOF + REALSXP) {
listoftype *q;
int j, end;
double *p;
q= PLIST(i);
p = q->p[r];
end = q->nrow[r] * q->ncol[r];
for (j=0; j<end; j++)
if (ISNAN(p[j])) ERR("no NAs allowed in regression ");
v = 0.0; // dummy
} else if (type[i] == CLOSXP) {
v = 0.0; //dummy
} else if (type[i] == LANGSXP) {
v = 0.0; //dummy
} else {
v = RF_NA;
BUG;
}
isnan[*NAs] = ISNAN(v) && !ISNA(v);
if (ISNAN(v)) { // entgegen Arith.h gibt ISNA nur NA an !!
if (isRandom(cov)){
NotProgrammedYet("NA in random functions");
}
if (printing > 0) {
if (nv>1 || (c>0 && printing > 1)) PRINTF(", "); else leer(depth+1);
}
if (isDollar(cov)) {
// shortD partial name for R level
cov_model *next = cov->sub[0];
while(isDollar(next) || isNatsc(next)) next = next->sub[0];// ok
if (covzaehler[next->nr] == 0) { // next wurde noch nicht
// untersucht, somit ist covzaehler[next->nr] um 1 niedriger
// als covzaehler[cov->nr] !
strcopyN(shortD, NICK(next) + 2, SHORTlen);
} else {
char dummy[255];
strcopyN(dummy, NICK(next) + 2, SHORTlen-1);
sprintf(shortD, "%s%d", dummy, covzaehler[next->nr]+1);
// print("$$ > %d:%s %d %s\n", next->nr, NICK(to),
// covzaehler[next->nr], shortD);
}
// assert(DANISO == DSCALE + 1);
if (i==DVAR) {
cov_model *call = cov->calling;
if (call == NULL) BUG;
if (!is_top(call) && call->nr == MIXEDEFFECT) {
sorts[*NAs] = !is_top(call->calling) &&
(call->calling->nr == PLUS || call->calling->nr == SELECT)
&& is_top(call->calling->calling) ? MIXEDVAR : ANYPARAM;
} else {
while (!is_top(call) && (call->nr == PLUS || call->nr==SELECT))
call = call->calling;
if (call == NULL) BUG;
sorts[*NAs] = is_top(call)
? (next->nr == NUGGET ? NUGGETVAR : VARPARAM)
: ANYPARAM;
}
if (no_variance && sorts[*NAs] <= SIGNEDSDPARAM)
sorts[*NAs] = ANYPARAM;
sprintf(names[*NAs],
sorts[*NAs] <= SIGNEDSDPARAM || sorts[*NAs] == NUGGETVAR
|| sorts[*NAs]==MIXEDVAR ? "%s.var" : "%s.vx",
shortD); // for R level only
} else {
if (i==DSCALE) {
// lastmem = mem[*NAs]; // used for all diagonal elements of
// aniso
sorts[*NAs] = SCALEPARAM;
sprintf(names[*NAs], "%s.s", shortD);// for R level only
} else {
assert(i == DANISO);
// no lastmem here !
if (cov->q != NULL && cov->q[0] != 0.0) {
// print("sdfasdf\n");
ERR("naturalscaling only allowed for isotropic setting");
}
if (r==c) {
sorts[*NAs] = DIAGPARAM;
sprintf(names[*NAs], "%s.d.%d", shortD, r);// R level only
} else {
sorts[*NAs] = ANISOPARAM;
sprintf(names[*NAs], "%s.d.%d.%d", shortD, r, c);// R level
}
}
}
} else {
// standard setting !
if (cov->nr==MIXEDEFFECT || cov->nr==TREND) {
// || cov->nr==MLEMIXEDEFFECT)
if (excludetrends) continue;
assert(type[i] == REALSXP);
sorts[*NAs] = cov->nr==TREND ? TRENDPARAM : MIXEDPARAM;
} else {
if (type[i] == INTSXP)
sorts[*NAs] = INTEGERPARAM;
else {
// r)ow, c)olumn; i:kappa, cov->nr, C
sorts[*NAs] = C->paramtype(i, r, c); // ANYPARAM;
if (sorts[*NAs] == IGNOREPARAM ||
sorts[*NAs] == DONOTRETURNPARAM) continue;
// print("%s k=%d no_var=%d, %d \n", C->name, i,
// no_variance, sorts[*NAs]);
if (no_variance && sorts[*NAs] <= SIGNEDSDPARAM)
sorts[*NAs] = ANYPARAM;
}
}
char kappashort[255];
strcopyN(kappashort, CC->kappanames[i], SHORTlen);
sprintf(names[*NAs], "%s.%s", shortname, kappashort);
if (*NAs > 0 && 0==
strncmp(names[*NAs], names[*NAs-1], strlen(names[*NAs]))){
if (namenr == 1) {
sprintf(names[*NAs-1], "%s.%s.%d",
shortname, kappashort, namenr);
}
namenr++;
sprintf(names[*NAs], "%s.%s.%d", shortname, kappashort,namenr);
} else namenr=1;
}
if (printing > 0) {
if (printing <= 2) PRINTF("%d",*NAs + 1);
else PRINTF("!%d", *NAs + 1);
}
// print("%d %ld\n", *NAs, mem[*NAs]);
(*NAs)++;
nv++;
} else { // not is.na(v)
// kein NA an der Position; somit nur Anzeige
if (printing > 1) {
if (c>0) PRINTF(", "); else leer(depth+1);
if (printing <= 2) PRINTF("*");
else {
if (type[i]== REALSXP) PRINTF("%6.2e", v);
else PRINTF("%5d", (int) v);
}
}
} // else kein NA
} //c
if ((printing > 0 && nv > 0) || (printing > 1 && ncol[i] > 0))
PRINTF("\n");
} // r
} // i
cov_model *sub;
for (i=0; i<MAXSUB; i++) {
sub = cov->sub[i];
if (sub != NULL) {
if (printing > 0) {
leer(depth);
PRINTF("%s = ", C->subnames[i]);
}
GetNAPosition(sub, NAs, mem, elmnts, mem_elmnts,
names, sorts, isnan,
covModels,
covzaehler, allowforintegerNA,
SHORTlen, printing, depth,
C->paramtype(-i-1, 0, 0) != VARPARAM,
excludetrends );
}
}
}
/* wird (derzeit) nicht verwenden !! */
SEXP GetNAPositions(SEXP model_reg, SEXP model, SEXP spatialdim, SEXP Time,
SEXP xdimOZ, SEXP integerNA, SEXP Print) {
int i, NAs, elmnts, covzaehler[MAXNRCOVFCTS];
naptr_type mem;
elptr_type mem_elmnts;
covptr_type covModels;
sortsofparam sorts[MAX_NA];
bool isnan[MAX_NA];
NAname_type names;
bool skipchecks = GLOBAL.general.skipchecks;
GLOBAL.general.skipchecks = true;
CheckModelInternal(model, ZERO, ZERO, ZERO, INTEGER(spatialdim)[0],
INTEGER(xdimOZ)[0], 1, 1, false, false,
LOGICAL(Time)[0],
KEY + INTEGER(model_reg)[0]);
sprintf(ERROR_LOC, "getting positions with NA: ");
// print("global=%d\n", GLOBAL.fit.lengthshortname);
GLOBAL.general.skipchecks = skipchecks;
NAs = 0;
for (i=0; i<MAXNRCOVFCTS; covzaehler[i++]=0);
GetNAPosition(KEY[MODEL_USER], &NAs, mem, &elmnts, mem_elmnts,
names, sorts, isnan,
covModels,
covzaehler,
INTEGER(integerNA)[0],
GLOBAL.fit.lengthshortname,
INTEGER(Print)[0],
0, false, true);
SEXP ans;
PROTECT (ans = allocVector(INTSXP, 1));
INTEGER(ans)[0] = NAs;
UNPROTECT(1);
return ans;
}
int countnas(cov_model *cov, int level) {
int i, r, count, endfor,
*nrow = cov->nrow,
*ncol = cov->ncol;
cov_fct *C = CovList + cov->nr; // nicht gatternr
SEXPTYPE *type = C->kappatype;
if (cov->nr == MIXEDEFFECT && level==0) { // || cov->nr == MLEMIXEDEFFECT )
if (cov->nrow[MIXED_BETA] > 0) return 0; // b is given
}
if (cov->nr == TREND && level==0) return 0;
count= 0;
for (i=0; i<C->kappas; i++) {
if (nrow[i] == 0 || ncol[i] == 0 || C->paramtype(i, 0, 0) == IGNOREPARAM
|| C->paramtype(i, 0, 0) == DONOTRETURNPARAM )
continue;
endfor = nrow[i] * ncol[i];
if (type[i] == REALSXP) {
double *p = P(i);
for (r=0; r<endfor; r++) if (ISNAN(p[r])) count++;
} else if (type[i] == INTSXP) {
int *p = PINT(i);
for (r=0; r<endfor; r++) if (p[r] == NA_INTEGER) count++;
} else if (type[i] == LISTOF + REALSXP) {
continue; // no NAs allowed
} else BUG;
}
cov_model *sub;
for (i=0; i<MAXSUB; i++) {
sub = cov->sub[i];
if (sub == NULL) continue;
count += countnas(sub, level + 1);
}
return count;
}
void Take21internal(cov_model *cov, cov_model *cov_bound,
double **bounds_pointer, int *NBOUNDS) {
/* determine the bounds for the parameters that
are to be estimated
*/
int i, c, r,
nv=0, // zaehlt NAs
*nrow = cov->nrow,
*ncol = cov->ncol,
*nrow2 = cov_bound->nrow,
*ncol2 = cov_bound->ncol;
cov_fct *C = CovList + cov->nr; // nicht gatternr
SEXPTYPE *type = C->kappatype;
if (strcmp(Nick(cov), Nick(cov_bound)) != 0) {
// print("%s %s\n", CovList[cov->nr].nick, CovList[cov_bound->nr].nick);
ERR("models do not match.");
}
// print("\n\n\n ======================= \n");
// print("%s %d %s\n", C->name, C->kappas, CovList[cov_bound->nr].name);
for (i=0; i<C->kappas; i++) {
// print("i=%d %s %s alt=%s\n", i, C->name, C->kappanames[i],
//CovList[cov_bound->nr].name);
if (C->kappatype[i] >= LISTOF ||
C->paramtype(i, 0, 0) == IGNOREPARAM ||
C->paramtype(i, 0, 0) == DONOTRETURNPARAM) continue;
if (nrow[i] != nrow2[i] || ncol[i] != ncol2[i]) {
PRINTF("%s i: %d, nrow1=%d, nrow2=%d, ncol1=%d, ncol2=%d\n",
C->name, i, nrow[i], nrow2[i], ncol[i], ncol2[i]);
ERR("lower/upper/user does not fit the model (size of matrix)");
}
for (r=0; r<nrow[i]; r++) {
for (c=0; c<ncol[i]; c++) {
double v=RF_NA,
w=RF_NA; // value in aktuellem parameter
int idx = c * nrow[i] + r;
// print("%d %d idx=%d %d %d %d nv=%d %d\n", r,c, idx, type[i], REALSXP, INTSXP,
// nv, *NBOUNDS);
if (type[i] == REALSXP) {
v = P(i)[idx];
w = PARAM(cov_bound, i)[idx];
}
else if (type[i] == INTSXP) {
v = PINT(i)[idx] == NA_INTEGER ? RF_NA : (double) PINT(i)[idx];
w = PARAMINT(cov_bound, i)[idx] == NA_INTEGER
? RF_NA : (double) PARAMINT(cov_bound, i)[idx];
}
if (ISNAN(v)) { // entgegen Arith.h gibt ISNA nur NA an !!
if ((!isDollar(cov) ||
i == DVAR ||
(i== DSCALE && cov->q == NULL) || // ! natscaling
i == DANISO) &&
(cov->nr != MIXEDEFFECT && cov->nr != TREND)
) // aniso ?? ABfrage OK ??
{
// print("%s %s, r=%d, c=%d: %d <? %d\n",
// C->name, C->kappanames[i], r, c, nv, *NBOUNDS);
if (nv >= *NBOUNDS) {
PRINTF("%s %s, r=%d, c=%d: %d >= %d\n",
C->name, C->kappanames[i], r, c, nv, *NBOUNDS);
ERR("lower/upper/user does not fit the model (number parameters)");
}
(*bounds_pointer)[nv] = w;
nv++;
}
} //ISNA
} //c
} // r
} // i
*NBOUNDS = *NBOUNDS - nv;
(*bounds_pointer) += nv;
for (i=0; i<MAXSUB; i++) {
if (cov->sub[i] != NULL) {
Take21internal(cov->sub[i], cov_bound->sub[i], bounds_pointer, NBOUNDS);
}
}
}
SEXP Take2ndAtNaOf1st(SEXP model_reg, SEXP model, SEXP model_bound,
SEXP spatialdim, SEXP Time, SEXP xdimOZ,
SEXP nbounds, SEXP skipchecks) {
// MAXMLEDIM
double *bounds_pointer;
int m,
NBOUNDS_INT =INTEGER(nbounds)[0],
*NBOUNDS= &NBOUNDS_INT,
nr[2] = {INTEGER(model_reg)[0], MODEL_BOUNDS};
SEXP bounds,
models[2] = {model, model_bound};
bool oldskipchecks = GLOBAL.general.skipchecks;
if (nr[0] == nr[1]) error("do not use register 'model bounds'");
NAOK_RANGE = true;
if (LOGICAL(skipchecks)[0]) GLOBAL.general.skipchecks = true;
for (m=1; m>=0; m--) { // 2er Schleife !!
// assert(m==1);
CheckModelInternal(models[m], ZERO, ZERO, ZERO, INTEGER(spatialdim)[0],
INTEGER(xdimOZ)[0], 1, 1, false, false,
LOGICAL(Time)[0],
KEY + nr[m]);
GLOBAL.general.skipchecks = oldskipchecks;
}
NAOK_RANGE = false;
PROTECT(bounds = allocVector(REALSXP, *NBOUNDS));
bounds_pointer = NUMERIC_POINTER(bounds);
// PMI(KEY[nr[0]]);
// PMI(KEY[nr[1]]);
Take21internal(KEY[nr[0]], KEY[nr[1]], &bounds_pointer, NBOUNDS);
if (*NBOUNDS != 0) ERR("lower/upper does not fit to model");
UNPROTECT(1);
return bounds;
}
void GetNARanges(cov_model *cov, cov_model *min, cov_model *max,
double *minpile, double *maxpile, int *NAs) {
/*
determine the ranges of the parameters to be estimated
*/
int i, r,
*nrow = cov->nrow,
*ncol = cov->ncol;
cov_fct *C = CovList + cov->nr; // nicht gatternr
SEXPTYPE *type = C->kappatype;
double v, dmin, dmax;
// print("\ngetnaranges\n");
for (i=0; i<C->kappas; i++) {
int end = nrow[i] * ncol[i];
if (end == 0) continue;
if (type[i] == REALSXP) {
dmin = PARAM0(min, i);
dmax = PARAM0(max, i);
} else if (type[i] == INTSXP) {
dmin = PARAM0INT(min, i) == NA_INTEGER
? RF_NA : (double) PARAM0INT(min, i);
dmax = PARAM0INT(max, i) == NA_INTEGER
? RF_NA : (double) PARAM0INT(max, i);
} else if (type[i] == LISTOF + REALSXP) {
dmin = PARAM0(min, i);
dmax = PARAM0(max, i);
} else if (type[i] == CLOSXP) {
dmin = 0.0;
dmax = 0.0;
} else if (type[i] == LANGSXP) {
dmin = 0.0;
dmax = 0.0;
} else {
BUG;
dmin = dmax = RF_NA;
}
for (r=0; r<end; r++) {
v = RF_NA;
if (type[i] == REALSXP) {
v = P(i)[r];
}
else if (type[i] == INTSXP) {
v = PINT(i)[r] == NA_INTEGER ? RF_NA : (double) PINT(i)[r];
} else if (type[i] == LISTOF + REALSXP) {
continue; // !!!!!!!!!!!
} else if (type[i] == CLOSXP) {
continue; // !!!!!!!!!!!
} else if (type[i] == LANGSXP) {
continue; // !!!!!!!!!!!
} else BUG;
if (ISNAN(v) && C->paramtype(i, 0, 0) != IGNOREPARAM
&& C->paramtype(i, 0, 0) != DONOTRETURNPARAM
&& cov->nr!=MIXEDEFFECT && cov->nr!=TREND) {// cov->nr!=MLEMIXEDEFFECT
if (!isDollar(cov) || (i!=DALEFT && i!=DPROJ)) {
minpile[*NAs] = dmin;
maxpile[*NAs] = dmax;
// print("%s %d %f %f\n", C->kappanames[i],r, dmin, dmax);
(*NAs)++;
} else {
// maybe NAs and internpile should be set here
continue;
}
} // isna
} // r
} // kappas
for (i=0; i<MAXSUB; i++) {
if (cov->sub[i] != NULL) {
GetNARanges(cov->sub[i],
min->sub[i], max->sub[i],
minpile, maxpile,
NAs);
}
}
}
int check_recursive_range(cov_model *cov, bool NAOK) {
/*
called by SetAndGetModelInfo
*/
int i, err;
if ((err = check_within_range(cov, NAOK)) != NOERROR) return err;
for (i=0; i<MAXSUB; i++) {
if (cov->sub[i] != NULL) {
if ((err = check_recursive_range(cov->sub[i], NAOK)) != NOERROR)
return err;
}
}
return NOERROR;
}
int CheckEffect(cov_model *cov) {
int i,j,end, nr;
bool na_var = false;
double *p;
cov_model *next;
if (cov->nr == MIXEDEFFECT) {
// cov->nr = MLEMIXEDEFFECT;
if (cov->nsub == 0) {
return (cov->nrow[MIXED_BETA] > 0 && ISNAN(P0(MIXED_BETA)))
? fixedeffect : deteffect;
}
next = cov->sub[0];
if (isDollar(next)) {
if (next->ncol[DVAR] == 1 && next->nrow[DVAR] == 1) {
na_var = ISNAN(PARAM0(next, DVAR));
}
for (i=0; i<=DMAX; i++) {
if (i!=DVAR) {
end = next->ncol[i] * next->nrow[i];
p = PARAM(next, i);
for (j=0; j<end; j++) {
if (ISNAN(p[j])) {
return next->nr == CONSTANT ? eff_error :
na_var ? spvareffect : spaceeffect;
// NA in e.g. scale for constant matrix -- does not make sense
}
}
}
}
next = next->sub[0]; // ok
}
int xx = next->nr == CONSTANT
? (cov->nrow[CONSTANT_M] > cov->ncol[CONSTANT_M]
? (na_var ? rvareffect : randomeffect)
: (na_var ? lvareffect : largeeffect)
)
: (na_var ? spvareffect : spaceeffect);
if (xx == spvareffect || xx == spaceeffect) {
BUG; //APMI(next);
}
return next->nr == CONSTANT
? (cov->nrow[CONSTANT_M] > cov->ncol[CONSTANT_M]
? (na_var ? rvareffect : randomeffect)
: (na_var ? lvareffect : largeeffect)
)
: (na_var ? spvareffect : spaceeffect);
}
if (cov->nr == TREND) {
int trend,
effect = eff_error;
bool isna ;
for (j=0; j<2; j++) {
trend = (j==0) ? TREND_MEAN : TREND_LINEAR;
// print("trend %d %d\n", j, trend);
if ( (nr = cov->nrow[trend] * cov->ncol[trend]) > 0) {
p = P(trend);
isna = ISNAN(p[0]);
if ((effect != eff_error) && ((effect == fixedtrend) xor isna))
SERR1("do not mix deterministic effect with fixed effects in '%s'",
NICK(cov));
for (i = 1; i<nr; i++) {
if (ISNAN(p[i]) xor isna)
SERR("mu and linear trend: all coefficient must be deterministic or all must be estimated");
}
effect = isna ? fixedtrend : dettrend;
}
}
for (j=0; j<2; j++) {
int param;
trend = (j==0) ? TREND_POLY : TREND_FCT;
//print("trend %d\n", trend);
param = (j==0) ? TREND_PARAM_POLY : TREND_PARAM_FCT;
if (cov->nrow[trend] > 0) {
if (effect != eff_error)
SERR("polynomials and free functions in trend may not be mixed with other trend definitions. Please use a sum of trends.");
nr = cov->nrow[param] * cov->ncol[param];
if (nr > 0) {
p = P(param);
isna = ISNAN(p[0]);
for (i = 1; i<nr; i++) {
if (ISNAN(p[i]) xor isna)
SERR("the coefficients in trend must be all deterministic or all coefficient are estimated");
}
effect = isna ? fixedtrend : dettrend;
} else effect = fixedtrend;
}
}
return effect;
}
cov_model *sub = cov;
bool Simple = true;
if (isDollar(sub)) {
Simple = PARAMisNULL(sub, DPROJ) && PARAMisNULL(sub, DANISO) &&
PARAMisNULL(sub, DALEFT);
sub = sub->sub[0];
}
if (isNatsc(sub)) sub = sub->sub[0];
cov_fct *C = CovList + sub->nr; // nicht gatternr
if (C->maxsub == 0) {
if (isPosDef(C->Type) &&
C->domain == XONLY && C->isotropy == ISOTROPIC &&
(C->vdim == 1 || C->cov==nugget) && Simple) {
return simple;
}
return primitive;
}
return remaining;
}
static naptr_type MEMORY[MODEL_MAX+1];
static covptr_type MEM_COVMODELS[MODEL_MAX+1];
static elptr_type MEMORY_ELMNTS[MODEL_MAX+1];
int MEM_NAS[MODEL_MAX+1], MEM_ELMNTS[MODEL_MAX+1];
SEXP SetAndGetModelInfo(SEXP model_reg, SEXP model, SEXP spatialdim,
SEXP distances, SEXP ygiven,
SEXP Time,
SEXP xdimOZ, SEXP shortlen, SEXP allowforintegerNA,
SEXP excludetrend) {
// ex MLEGetNAPos
sortsofparam MEM_SORTS[MAX_NA];
bool anyeffect, MEM_ISNAN[MAX_NA];
/*
basic set up of covariance model, determination of the NA's
and the corresponding ranges.
*/
cov_model *cov, *key,
*min=NULL, *max=NULL,
*pmin=NULL, *pmax=NULL,
*openmin=NULL, *openmax=NULL;
// bool skipchecks = GLOBAL.general.skipchecks;
NAname_type names;
double mle_min[MAX_NA], mle_max[MAX_NA], mle_pmin[MAX_NA], mle_pmax[MAX_NA],
mle_openmin[MAX_NA], mle_openmax[MAX_NA];
int i, covzaehler[MAXNRCOVFCTS], effect[MAXSUB], nas[MAXSUB],
newxdim, neffect, NAs,
err = NOERROR,
modelnr = INTEGER(model_reg)[0];
const char *colnames[8] =
{"pmin", "pmax", "type", "is.nan", "min", "max", "openmin", "openmax"};
SEXP trans_inv, isotropic, Sxdim,
matrix, nameAns, nameMatrix, RownameMatrix,
ans=NULL;
#define total 7
PROTECT(trans_inv=allocVector(LGLSXP, 1));
LOGICAL(trans_inv)[0] = false;
PROTECT(isotropic=allocVector(LGLSXP, 1));
LOGICAL(isotropic)[0] = false;
NAOK_RANGE = true;
CheckModelInternal(model, ZERO, ZERO, ZERO,
INTEGER(spatialdim)[0], INTEGER(xdimOZ)[0],
1, LOGICAL(ygiven)[0] ? 1 : 0, // lx, ly
false, //grid
LOGICAL(distances)[0], //distances
LOGICAL(Time)[0],
KEY + modelnr);
// APMI(KEY[modelnr]); //, "setandget");
cov = key = KEY[modelnr];
if (isInterface(cov)) {
cov = cov->key == NULL ? cov->sub[0] : cov->key;
}
if (key->prevloc->timespacedim > MAXMLEDIM)
ERR("dimension too high in MLE");
NAOK_RANGE = false;
if (cov->pref[Nothing] == PREF_NONE) {
err = ERRORINVALIDMODEL;
goto ErrorHandling;
}
// print("here returnx tsdim %d %d %d\n",
// INTEGER(tsdim)[0], LOGICAL(domain)[0], LOGICAL(isotropic)[0]);
// print("here\n");
// PMI(cov);
newxdim = INTEGER(xdimOZ)[0];
if (cov->domprev == XONLY && isPosDef(cov->typus)) {
LOGICAL(trans_inv)[0] = true;
if (cov->isoown==ISOTROPIC) {
LOGICAL(isotropic)[0] = true;
newxdim = 1;
// removeOnly(KEY + modelnr); nicht wegnehmen
// da derzeit negative distanczen in GLOBAL.CovMatrixMult auftreten -- obsolete ?!
}
}
// print("SD here returnx tsdim %d\n", INTEGER(tsdim)[0]);
// GLOBAL.general.skipchecks = skipchecks;
check_recursive_range(cov, true);
// print("AS here returnx tsdim %d\n", INTEGER(tsdim)[0]);
if ((err = get_ranges(cov, &min, &max, &pmin, &pmax, &openmin, &openmax))
!= NOERROR) goto ErrorHandling;
// print("hier %d\n", INTEGER(shortlen)[0]);
MEM_ELMNTS[modelnr] = MEM_NAS[modelnr] = 0;
for (i=0; i<MAXNRCOVFCTS; covzaehler[i++]=0);
// !!
GetNAPosition(cov, MEM_NAS + modelnr, MEMORY[modelnr],
MEM_ELMNTS + modelnr, MEMORY_ELMNTS[modelnr],
names, MEM_SORTS, MEM_ISNAN,
MEM_COVMODELS[modelnr],
covzaehler,
INTEGER(allowforintegerNA)[0],
INTEGER(shortlen)[0],
(PL >= PL_COV_STRUCTURE) + (PL >=PL_DETAILS) +
(PL >= PL_SUBDETAILS),
0, false, LOGICAL(excludetrend)[0]);
// print("XX modelnr=%d\n", modelnr);
NAs = 0; GetNARanges(cov, min, max, mle_min, mle_max, &NAs);
//print("XXz modelnr=%d\n", modelnr);
NAs = 0; GetNARanges(cov, pmin, pmax, mle_pmin, mle_pmax, &NAs);
//print("XX-------- amodelnr=%d\n", modelnr);
NAs = 0; GetNARanges(cov, openmin, openmax, mle_openmin, mle_openmax, &NAs);
anyeffect=false;
if (cov->nr == PLUS || cov->nr == SELECT) {
for (i=0; i<cov->nsub; i++) {
// print("checkeffect: %d %d\n", i, cov->nsub);
effect[i] = CheckEffect(cov->sub[i]);
nas[i] = countnas(cov->sub[i], 0);
// print("i=%d nsub=%d eff=%d remain=%d nas=%d\n", i, cov->nsub, effect[i], remaining, nas[i]);
if (effect[i] == eff_error) GERR("scaling parameter appears with constant matrix in the mixed effect part");
anyeffect |= (effect[i] != remaining && effect[i] != primitive);
}
} else {
effect[0] = CheckEffect(cov);
nas[0] = countnas(cov, 0);
}
anyeffect = true; // always bring back effects!!
if (anyeffect) {
if (cov->nr == PLUS || cov->nr == SELECT)
neffect = cov->nsub;
else neffect = 1;
} else neffect = 0;
// print("There returnx tsdim %d\n", INTEGER(tsdim)[0]);
PROTECT(Sxdim=allocVector(INTSXP, 1));
INTEGER(Sxdim)[0] = newxdim;
// print("hYere returnx tsdim %d\n", INTEGER(tsdim)[0]);
PROTECT(matrix = allocMatrix(REALSXP, MEM_NAS[modelnr], 8));
PROTECT(RownameMatrix = allocVector(STRSXP, MEM_NAS[modelnr]));
PROTECT(nameMatrix = allocVector(VECSXP, 2));
for (i=0; i<MEM_NAS[modelnr]; i++) {
REAL(matrix)[i] = mle_pmin[i];
REAL(matrix)[i + MEM_NAS[modelnr]] = mle_pmax[i];
REAL(matrix)[i + 2 * MEM_NAS[modelnr]] = MEM_SORTS[i];
REAL(matrix)[i + 3 * MEM_NAS[modelnr]] = (int) MEM_ISNAN[i];
REAL(matrix)[i + 4 * MEM_NAS[modelnr]] = mle_min[i];
REAL(matrix)[i + 5 * MEM_NAS[modelnr]] = mle_max[i];
REAL(matrix)[i + 6 * MEM_NAS[modelnr]] = mle_openmin[i];
REAL(matrix)[i + 7 * MEM_NAS[modelnr]] = mle_openmax[i];
SET_STRING_ELT(RownameMatrix, i, mkChar(names[i]));
// print("i=%d %s %e %e\n", i, names[i], mle_pmin[i], mle_pmax[i]);
}
SET_VECTOR_ELT(nameMatrix, 0, RownameMatrix);
SET_VECTOR_ELT(nameMatrix, 1, Char(colnames, 8));
setAttrib(matrix, R_DimNamesSymbol, nameMatrix);
PROTECT(ans = allocVector(VECSXP, total));
PROTECT(nameAns = allocVector(STRSXP, total));
i = 0;
SET_STRING_ELT(nameAns, i, mkChar("minmax"));
SET_VECTOR_ELT(ans, i++, matrix);
SET_STRING_ELT(nameAns, i, mkChar("trans.inv")); // !!
SET_VECTOR_ELT(ans, i++, trans_inv);
SET_STRING_ELT(nameAns, i, mkChar("isotropic"));
SET_VECTOR_ELT(ans, i++, isotropic);
SET_STRING_ELT(nameAns, i, mkChar("effect"));
SET_VECTOR_ELT(ans, i++, Int(effect, neffect, MAXINT));
SET_STRING_ELT(nameAns, i, mkChar("NAs"));
SET_VECTOR_ELT(ans, i++, Int(nas, neffect, MAXINT));
SET_STRING_ELT(nameAns, i, mkChar("xdimOZ"));
SET_VECTOR_ELT(ans, i++, Sxdim);
SET_STRING_ELT(nameAns, i, mkChar("matrix.indep.of.x"));
SET_VECTOR_ELT(ans, i++, ScalarLogical(cov->matrix_indep_of_x));
setAttrib(ans, R_NamesSymbol, nameAns);
assert(i==total);
UNPROTECT(8);
// print(" J here returnx tsdim %d\n", INTEGER(tsdim)[0]);
ErrorHandling:
if (min != NULL) COV_DELETE(&min);
if (max != NULL) COV_DELETE(&max);
if (pmin != NULL) COV_DELETE(&pmin);
if (pmax != NULL) COV_DELETE(&pmax);
if (openmin != NULL) COV_DELETE(&openmin);
if (openmax != NULL) COV_DELETE(&openmax);
if (err != NOERROR) XERR(err);
return ans;
}
void expliciteDollarMLE(int* ModelNr, double *values) { //
// userinterfaces.cc
int i, un,
modelnr = *ModelNr,
NAs = MEM_NAS[modelnr];
// first get the naturalscaling values and devide the preceeding
// scale model by this value
if (NS==NATSCALE_MLE) {
iexplDollar(KEY[modelnr], true);
}
// Then get the values out of the model
for (un=i=0; un<NAs; i++) {
values[un++] = MEMORY[modelnr][i][0];
MEMORY[modelnr][i][0] = RF_NA;
}
}
void PutValuesAtNAintern(int *reg, double *values, bool init){
int i, un,
NAs = MEM_NAS[*reg];
cov_fct *C = NULL;
cov_model *cov = NULL;
gen_storage s;
STORAGE_NULL(&s);
s.check = false;
// set ordinary parameters for all (sub)models
// cov_model *cov = KEY[*reg];
//cov = cov->sub[0];
// print("%s %d\n", NICK(cov), *reg);
///assert(isDollar(cov));
//print("%ld %f %ld %f\n", cov->p[DVAR], cov->p[DVAR][0],
// cov->p[DSCALE], cov->p[DSCALE][0]);
// printf("putvaluesatna\n");
for (un=i=0; i<NAs; i++) {
// print("reg=%d i=%d %d %ld %f\n", *reg, i, NAs, MEMORY[*reg][i], values[un]);
// print("mem=%ld %f\n", (long int) MEMORY[*reg][i], values[un]) ;
//if (*reg < 0 || *reg > MODEL_MAX) XERR(ERRORREGISTER);
MEMORY[*reg][i][0] = values[un++];
}
if (init)
for (i=0; i<NAs; i++) {
cov = MEM_COVMODELS[*reg][i];
C = CovList + cov->nr;
if (i==0 || cov != MEM_COVMODELS[*reg][i-1]) {
if (!isDummyInit(C->Init)) {
// print("i=%d %s initalising (%f)\n", i, NICK(MEM_COVMODELS[*reg][i]), MEMORY[*reg][i][0]);
C->Init(cov, &s);
//print("i=%d %s initalised (%f)\n", i, NICK(MEM_COVMODELS[*reg][i]),
// MEMORY[*reg][i][0]);
}
}
//
}
//printf("done\n");
int one = 1;
setListElements(reg, &one, &one, &one);
}
void PutValuesAtNA(int *reg, double *values){
PutValuesAtNAintern(reg, values, true);
}
void PutValuesAtNAnoInit(int *reg, double *values){
PutValuesAtNAintern(reg, values, false);
}
void setListElements(int *reg, int *i, int *k, int* len_k) {
int len = *len_k;
if (*reg < 0 || *reg > MODEL_MAX) XERR(ERRORREGISTER);
cov_model *cov = KEY[*reg];
if (cov==NULL) ERR("register is not initialised by 'RFfit'");
if (isInterface(cov)) {
cov = cov->key == NULL ? cov->sub[0] : cov->key;
}
if (cov->nr == SELECT) {
int j;
if (len != cov->nrow[SELECT_SUBNR]) {
PFREE(SELECT_SUBNR);
PALLOC(SELECT_SUBNR, len, 1);
}
for (j=0; j<len; j++) {
PINT(SELECT_SUBNR)[j] = k[j] - 1;
}
}
int j,
iM1 = *i - 1,
nr = MEM_ELMNTS[*reg];
for (j=0; j<nr; j++) {
MEMORY_ELMNTS[*reg][j][0] = iM1;
// printf("setlistel %d %d %ld\n", j, iM1, MEMORY_ELMNTS[*reg][j]);
}
}