swh:1:snp:8c443ee102a7f3a9fe0e834b4caec2f704d4d442
Tip revision: e10243fbd4eb0cbeaf518e67fbc5b8ad44889954 authored by Martin Schlather on 12 December 2019, 13:40:13 UTC
version 3.3.7
version 3.3.7
Tip revision: e10243f
gauss.cc
/*
Authors
Martin Schlather, schlather@math.uni-mannheim.de
Definition of correlation functions and derivatives of hypermodels
Copyright (C) 2005 -- 2017 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.
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 <Rmath.h>
#include "questions.h"
#include "Processes.h"
#include "Coordinate_systems.h"
#include "variogramAndCo.h"
#include "shape.h"
void location_rules(model *cov, pref_type pref) {
// 1. rules that depend on the the locations and the user's preferences only,
// but not on the covariance model#
if (COVNR != GAUSSPROC && COVNR != BINARYPROC) BUG;
location_type *loc = Loc(cov);
usr_bool exactness = GLOBAL.general.exactness; //
unsigned int maxmem=500000000;
int i;
Methods Standard[Nothing] = {
CircEmbed, CircEmbedIntrinsic, CircEmbedCutoff, SpectralTBM, TBM,
Direct, Specific, Sequential, Trendproc, Average, Nugget, RandomCoin,
Hyperplane
};
for (i=0; i<Nothing; i++) {
pref[Standard[i]] = Nothing - i;
}
if (P0INT(GAUSSPROC_STATONLY) == (int) True) {
pref[CircEmbedIntrinsic] = LOC_PREF_NONE - 1;
}
if (exactness == True) {
pref[TBM] = pref[SpectralTBM] = pref[Average] = pref[RandomCoin] =
pref[Sequential] = pref[Hyperplane]
= LOC_PREF_NONE - 2;
}
if (loc->timespacedim == 1) pref[TBM] -= 2 * PREF_PENALTY;
if (loc->distances) {
if (loc->grid) BUG;
for (i=CircEmbed; i<Nothing; i++) pref[i] = (i==Direct) * LOC_PREF_NONE;
} else if (loc->grid) {
if (exactness != True &&
loc->totalpoints * (1 << loc->timespacedim) * sizeof(double) > maxmem){
pref[CircEmbed] -= PREF_PENALTY;
pref[CircEmbedIntrinsic] -= PREF_PENALTY;
pref[CircEmbedCutoff] -= PREF_PENALTY;
}
} else {
if (exactness == True){
pref[CircEmbed] = pref[CircEmbedIntrinsic] = pref[CircEmbedCutoff] = -3;
} else {
pref[CircEmbed] -= PREF_PENALTY;
pref[CircEmbedIntrinsic] = -3;
pref[CircEmbedCutoff] -= PREF_PENALTY;
}
if (!loc->Time) pref[Sequential] = LOC_PREF_NONE;
}
}
void mixed_rules(model *cov, pref_type locpref,
pref_type pref, int *order) {
assert(COVNR == GAUSSPROC);
model *sub = cov->sub[0];
location_type *loc = Loc(cov);
int i,
*covpref = sub->pref,
totalpref[Nothing],
best_dirct=GLOBAL.gauss.direct_bestvariables,
max_variab= MAX(GLOBAL.direct.maxvariables, GLOBAL_UTILS->solve.max_chol),
vdim = VDIM0;
for (i=0; i<Nothing; i++) {
totalpref[i] = i == Nugget ? PREF_NUGGET * (PREF_FACTOR + 1): PREF_BEST;
// printf("%d %d", totalpref[i], covpref[i]);
if (totalpref[i] > covpref[i]) totalpref[i] = covpref[i];
if (totalpref[i] > PREF_NONE && locpref[i] > LOC_PREF_NONE)
pref[i] = totalpref[i] * PREF_FACTOR + locpref[i];
else
pref[i] = locpref[i] <= LOC_PREF_NONE ? locpref[i] : LOC_PREF_NONE - 4;
// printf("i=%d %d [%d, %d]\n", i, pref[i], locpref[i], totalpref[i]);
}
int vdimtot = loc->totalpoints * vdim;
if (vdimtot > max_variab &&
(sub->finiterange == falsch || GLOBAL_UTILS->solve.sparse == False))
pref[Direct] = LOC_PREF_NONE - 0;
// printf("spam %10g %d %10g %d %d\n", GLOBAL_UTILS->solve.sparse, GLOBAL_UTILS->solve.sparse == false, sub->finiterange,sub->finiterange==false, pref[Direct]);
if (vdimtot <= best_dirct && totalpref[Direct] == PREF_BEST) {
pref[Direct] = (PREF_BEST + 1 + (int) (best_dirct>=256 && vdimtot<256)) *
PREF_FACTOR;
}
else if (pref[Direct] >= PREF_NONE && GLOBAL_UTILS->solve.sparse != True) {
double ratio = -0.1;
if (max_variab <= DIRECT_ORIG_MAXVAR)
ratio = (double) (vdimtot - best_dirct) / (double) max_variab;
ratio *= FABS(ratio);
//printf("pref[direct] = %d %d %d\n", pref[Direct],max_variab, DIRECT_ORIG_MAXVAR );
pref[Direct] = (int) POW((double) pref[Direct], 1.0 - ratio);
if (pref[Direct] < PREF_BEST)
pref[Direct] = sub->finiterange == wahr ? PREF_BEST : PREF_BEST / 2;
//printf("ratio =%f\n", ratio);
}
if (P0INT(GAUSSPROC_STATONLY) == NA_LOGICAL && isnowPosDef(cov)) {
pref[CircEmbedIntrinsic] = LOC_PREF_NONE - 6;
}
if (!isCartesian(OWNISO(0))) {
pref[CircEmbedIntrinsic] = pref[CircEmbed] = pref[CircEmbedCutoff] =
LOC_PREF_NONE - 7;
if (isAnySpherical(OWNISO(0)) && OWNTOTALXDIM < 3)
pref[Sequential] = LOC_PREF_NONE - 8;
}
Ext_orderingInt(pref, Nothing, 1, order);
}
bool NAequal(double X, double Y) {
return (ISNAN(X) && ISNAN(Y)) || X == Y;
}
int kappaBoxCoxParam(model *cov, int BC) {
int vdim_2 = VDIM0 * 2;
if (PisNULL(BC)) {
PALLOC(BC, 2, VDIM0);
if (GLOBAL.gauss.loggauss) {
for (int i=0; i<vdim_2; i++) P(BC)[i] = 0.0;
GLOBAL.gauss.loggauss = false;
} else {
for (int i=0; i<vdim_2; i++)
P(BC)[i] = GLOBAL.gauss.boxcox[i];
}
} else {
int total = cov->nrow[BC] * cov->ncol[BC];
if (total == 1) {
double _boxcox = P0(BC);
PFREE(BC);
PALLOC(BC, 2, 1);
P(BC)[0] = _boxcox;
P(BC)[1] = GLOBAL.gauss.boxcox[1];
total = 2;
}
if (total < vdim_2) SERR("too few parameters for boxcox given");
if (total % 2 != 0) SERR("number of parameters of boxcox must even");
if (cov->nrow[BC] > 2 && cov->ncol[BC] > 1) SERR("parameters of boxcox must be given as a matrix with 2 rows");
cov->ncol[BC] = total / 2;
cov->nrow[BC] = 2;
bool notok = false;
if (GLOBAL.gauss.loggauss) {
for (int i=0; i<VDIM0; i++) {
if ((notok = P(BC)[2*i] != RF_INF &&
(P(BC)[2*i] != 0.0 || P(BC)[2*i+1] != 0.0)))
break;
}
} else {
for (int i=0; i<VDIM0; i++) {
//printf("%d %10g %10g %10g %10g\n", i, GLOBAL.gauss.boxcox[2 * i], GLOBAL.gauss.boxcox[2 * i + 1], P(BC)[2 * i], P(BC)[2 * i +1]);:
if ((notok = (GLOBAL.gauss.boxcox[2 * i] != RF_INF &&
!NAequal(P(BC)[2 * i], GLOBAL.gauss.boxcox[2 * i])) ||
!NAequal(P(BC)[2 * i +1], GLOBAL.gauss.boxcox[2 * i + 1])))
break;
}
}
if (notok)
SERR("Box Cox transformation is given twice, locally and through 'RFoptions'");
}
for (int i=0; i<VDIM0; i++) {
GLOBAL.gauss.boxcox[2 * i] = RF_INF;
GLOBAL.gauss.boxcox[2 * i + 1] = 0.0;
}
RETURN_NOERROR;
}
void kappaGProc(int i, model *cov, int *nr, int *nc){
*nc = i == GAUSS_BOXCOX ? SIZE_NOT_DETERMINED : 1;
*nr = i == GAUSS_BOXCOX ? SIZE_NOT_DETERMINED
: i < DefList[COVNR].kappas ? 1 : -1;
}
int checkgaussprocess(model *cov) {
// printf("\nGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG\n");
ASSERT_ONESYSTEM;
model
*next=cov->sub[cov->sub[0] == NULL],
*key = cov->key;
//location_type *loc=Loc(cov);
int err,
xdim = OWNXDIM(0), // could differ from logicaldim in case of distances!
dim = OWNLOGDIM(0);
gauss_param *gp = &(GLOBAL.gauss);
assert((Loc(cov)->distances && xdim==1) || xdim == dim);
if (!hasAnyProcessFrame(cov) &&
!(hasInterfaceFrame(cov) && cov->calling != NULL &&
cov->calling->calling == NULL) &&
!hasAnyEvaluationFrame(cov)) {
ILLEGAL_FRAME;
}
kdefault(cov, GAUSSPROC_STATONLY, (int) gp->stationary_only);
if (MAX(GLOBAL.direct.maxvariables, GLOBAL_UTILS->solve.max_chol) <
gp->direct_bestvariables)
SERR("maximum variables less than bestvariables for direct method");
if ((err = checkkappas(cov, false)) != NOERROR) RETURN_ERR(err);
set_maxdim(OWN, 0, INFDIM);
if (key == NULL) {
if (isGaussMethod(next)) {// waere GaussMethodType
// wird verboten
SERR1("%.50s may not call a method", NICK(cov));
} else {
Types frame = hasAnyEvaluationFrame(cov) ? cov->frame :
EvaluationType; // sehr schwach, sonst kommst aber trend nicht durch
//GaussMethodType;
// printf("@@@@@@@@@@@@ %d\n", hasEvaluationFrame(cov) );
if ((err = CHECKPOS2VAR(next, SUBMODEL_DEP, frame, KERNEL)) != NOERROR){
// if (NEXTTYPE(0)!=TrendType) {APMI(cov);BUG} else printf("jetzt C HECK_GEN\n");
set_type(PREVSYSOF(next), 0, TrendType);
set_iso(PREVSYSOF(next), 0, OWNISO(0));
set_dom(PREVSYSOF(next), 0, XONLY);
if ((err = CHECK_GEN(next, SUBMODEL_DEP, // err mocht verwemdet
SUBMODEL_DEP, TrendType, false)) != NOERROR) {
RETURN_ERR(err); // previous error
}
// printf("jff etzt C HECK_GEN\n");
}
/*
if ((err = CHECKPD2ND(next, dim, xdim,
Sy mmetricOf(OWNISO(0)),// Jan 2015 S YMMETRIC,
SUBMODEL_DEP, frame))
!= NOERROR) {
if (CHECK(next, dim, dim, TrendType, XONLY, OWNISO(0), // err mocjt verwemdet
SUBMODEL_DEP, frame)) RETURN_ERR(err); // previous error
*/
}
} else {
if (PL >= PL_COV_STRUCTURE) {
PRINTF("checking key in gauss process ...\n");
}
Types frame = hasAnyEvaluationFrame(cov) ? cov->frame : GaussMethodType;
if ((err = CHECK(key, dim, xdim, ProcessType, XONLY, OWNISO(0),
SUBMODEL_DEP, frame)) != NOERROR) RETURN_ERR(err);
}
model *sub = cov->key==NULL ? next : key;
setbackward(cov, sub);
if ((err = kappaBoxCoxParam(cov, GAUSS_BOXCOX)) != NOERROR) RETURN_ERR(err);
if ((err = checkkappas(cov, true)) != NOERROR) RETURN_ERR(err);
RETURN_NOERROR;
}
void rangegaussprocess(model VARIABLE_IS_NOT_USED *cov, range_type *range) {
GAUSS_COMMON_RANGE;
booleanRange(GAUSSPROC_STATONLY);
}
int gauss_init_settings(model *cov) {
model
*next = cov->sub[cov->sub[0] == NULL],
*sub = cov->key == NULL ? next : cov->key;
double sigma, var, meanDsigma,
*mean = NULL,
*variance = NULL;
int w, v, nmP1,
err = NOERROR,
vdim = next->vdim[0],
vdimSq = vdim * vdim,
vdimP1 = vdim + 1;
if ((variance = (double*) MALLOC(sizeof(double) * vdimSq)) == NULL ||
(mean = (double*) CALLOC(vdim, sizeof(double))) == NULL
) goto ErrorHandling;
GetInternalMean(next, vdim, mean);
if (ISNAN(mean[0])) // GetInternalMean currently only allows ...
GERR("Mean equals NA. Did you try a multivariate model or an incomplete (mixed) model?");
// if (false && cov->key != NULL) { // sind diese Zeilen notwendig?
// err = CHECK(next, OWNLOGDIM(0), OWNXDIM(0), PosDefType, KERNEL,
// , cov->vdim, EvaluationType);
// if (err != NOERROR) goto ErrorHandling;
// }
// if (!next->checked) crash();
assert(next->checked);
if (isXonly(PREVSYSOF(next))) {
COV(ZERO(next), next, variance);
} else {
assert(isKernel(PREVSYSOF(next)));
//double *zero = ZERO(next);
// NONSTATCOV(zero, zero, next, variance);
for (v=0; v<vdimSq; variance[v++] = 0.0); // just a dummy
}
if (cov->q == NULL) QALLOC(vdim);
if ((err = alloc_mpp_M(cov, 2)) != NOERROR) goto ErrorHandling;
nmP1 = cov->mpp.moments + 1;
for (v=w=0; w<vdimSq; w+=vdimP1, v++) {
sigma = SQRT(variance[w]);
meanDsigma = sigma == 0 ? RF_INF : mean[v] / sigma;
//the following line is SQRT(2 pi s^2) int_0^infty x EXP(-0.5 (x-mu)^2/s^2)
var = sigma * INVSQRTTWOPI * EXP(-0.5 * meanDsigma * meanDsigma) +
mean[v] * /* correct next lines the 2nd factor is given */
pnorm(0.0, mean[v], sigma, false, false);
var = 1.0 / (var * var);
cov->q[v] = var;
// trend um factor var korrigieren in rfmodel
// eigentliches model um factor var korrigieren
// so dass ein Gausssches Zufallsfeld rauskommt, dessen Positivteil
// zumindest im Ursprung den Erwartungswert 1 hat.
cov->mpp.maxheights[v] =
GLOBAL.extreme.standardmax * sigma + ((mean[v]>0) ? mean[v] : 0);//approx!
int idx = v * nmP1;
cov->mpp.mM[idx + 0] = cov->mpp.mMplus[idx + 0] = 1.0;
cov->mpp.mMplus[idx + 1] =
sigma * INVSQRTTWOPI * EXP(-0.5 * mean[v] * mean[v])
+ mean[v] * pnorm(-mean[v], 0.0, 1.0, false, false);
//(2pi)^-0.5 int x EXP(-(x-m)/2s^2)
cov->mpp.mM[idx + 1] = 0.0;
cov->mpp.mM[idx + 2] = variance[w];
// todo: cov->mpp.mMplus[2] berechnen
}
ReturnOtherField(cov, sub);
ErrorHandling:
FREE(variance);
FREE(mean);
RETURN_ERR(err);
}
int struct_gaussmethod(model *cov, model **newmodel) {
// uebernimmt struct- und init-Aufgaben !!
model *next = cov->sub[0]; // nicht cov->sub[cov->sub != NULL]
SPRINTF(cov->base->error_loc, "simulation procedure for %.50s", NAME(next));
ASSERT_ONESYSTEM;
location_type *loc=Loc(cov);
int
nr = COVNR,
err = ERROROUTOFMETHODLIST,
xdim = OWNXDIM(0), // could differ from logicaldim in case of distances!
logicaldim = OWNLOGDIM(0);
assert((loc->distances && xdim==1) || xdim == logicaldim);
cov->fieldreturn = wahr;
ASSERT_NEWMODEL_NULL;
assert(cov!=NULL);
if ((logicaldim != PREVXDIM(0) || logicaldim != xdim) &&
(!loc->distances || PREVXDIM(0)!=1)) {
BUG; // RETURN_ERR(ERRORDIM);
}
if (next!= NULL && !isnowVariogram(next)) {
SERR("submodel not a covariance function");
}
if (cov->key != NULL) COV_DELETE(&(cov->key), cov);
if ((err = covcpy(&(cov->key), cov)) != NOERROR) RETURN_ERR(err); //!!cov
defn *C = DefList + nr;
set_nr(SYSOF(cov->key), C->Specific);
// cov!! nicht cov->key!! ??? Nein, sonst ist Coin u.ae. nicht gesetzt.
model
*firstmethod = cov->key; // z.B. coins_intern
int idx = firstmethod->sub[0] == NULL;
model
*sub = firstmethod->sub[idx]; // z.B. oberste fkt shape von coins
if ((err = CHECK_PASSTF(cov->key, // 11.1.19 cov WARUM ??
GaussMethodType, VDIM0,
// frame // 19.5.2013 auskommentiert und ersetzt durch
// WARUM ??
GaussMethodType)) != NOERROR && !isAnyDollar(sub)) {
RETURN_ERR(err);
}
// if ((err = CHECK( c o v, dim, xdim, GaussMethodType, OWNDOM(0),
// OWNISO(0), cov->vdim,
// // frame // 19.5.2013 auskommentiert und ersetzt durch
// AnyType // wegen spectral.cc
// )) != NOERROR) {
// RETURN_ERR(err);
// }
if (err == NOERROR) err = STRUCT(cov->key, NULL);
if (err != NOERROR && !isAnyDollar(sub)) RETURN_ERR(err);
model *dollar,
*lastmethod= equalsnowGaussMethod(sub) ? sub : firstmethod;
if (err != NOERROR) {
assert(isAnyDollar(sub));
dollar = cov->key = lastmethod->sub[idx]; // $ jetzt im cov->key
lastmethod->sub[idx] = dollar->sub[0]; // eigentl. Modell jetzt anhaengen
SET_CALLING(lastmethod->sub[idx], lastmethod);
dollar->sub[0] = firstmethod; // auf $ setzen
SET_CALLING(firstmethod, dollar);
SET_CALLING(dollar, cov); // $ mit cov verbinden
dollar->prevloc = cov->prevloc;
set_nr(SYSOF(dollar), DOLLAR_PROC);
Types frame =
nr == AVERAGE_USER || nr == RANDOMCOIN_USER
? PoissonGaussType
: GaussMethodType;
/// cov nicht cov->key !!! ???? OK. Da sonst cov u.U. nicht gesetzt
if ((err = CHECK_PASSTF(cov->key, // 11.1.19 cov WARUM ??
GaussMethodType, //?! 8.12.17
VDIM0, frame)) != NOERROR) {
RETURN_ERR(err);
}
if ((err = STRUCT(cov->key, NULL)) != NOERROR) { RETURN_ERR(err); }
}
Methods m;
for (m=CircEmbed; m <= Hyperplane; m = (Methods) ((int) (m + 1))) {
if (gaussmethod[(Methods) m] == COVNR) break;
}
cov->key->method = m;
RETURN_NOERROR;
}
int struct_gaussprocess(model *cov, model **newmodel) {
// uebernimmt struct- und init-Aufgaben !!
KEY_type *KT = cov->base;
model *next = cov->sub[cov->sub[0] == NULL];
SPRINTF(KT->error_loc, "simulation method for %.50s", NAME(next));
ASSERT_NEWMODEL_NULL;
assert(cov!=NULL);
if (hasAnyEvaluationFrame(cov)) {
if (VDIM0 > MAXGAUSSVDIM)
SERR2("The number of variables, %d, exceeds the maximum number %d",
VDIM0, MAXGAUSSVDIM);
return struct_gauss_logli(cov); //loglihood
} else if (!hasAnyProcessFrame(cov) &&
!(hasInterfaceFrame(cov) && cov->calling != NULL &&
cov->calling->calling == NULL)
) RETURN_ERR(ERRORFAILED);
cov->fieldreturn = wahr;
location_type *loc = Loc(cov);
pref_type locpref, pref;
int order[Nothing], i;
char dummy[100 * (int) Nothing];
int
xdim = OWNXDIM(0), // could differ from tsdgetim in case of distances!
logicaldim = OWNLOGDIM(0);
assert((loc->distances && xdim==1) || xdim == logicaldim);
#define MAXFAILMSG 9
char FailureMsg[MAXFAILMSG][80] = {"unknown reason",
"total number of points > maximum value",
"non-domain model not allowed",
"not an exact method as required by user",
"locations not on a grid",
"denied by cov. model or by user's pref.",
"no analytic solution of Abelian integral",
"intrinsic fields denied by user",
"Unknown reason"
};
bool all_PREF_NONE;
Methods unimeth = Forbidden;
int
meth,
zaehler = 0,
nr = NEXTNR;
#define nsel 4
ASSERT_ONESYSTEM;
if( (logicaldim != PREVXDIM(0) || logicaldim != xdim) &&
(!loc->distances || PREVXDIM(0) !=1)) {
BUG; //RETURN_ERR(ERRORDIM);
}
if (cov->key != NULL) COV_DELETE(&(cov->key), cov);
if (!isnowVariogram(next) && !equalsnowTrend(next)) {
SERR("submodel must be a covariance function");
}
//////////////////////////////////////////////////////////////////////
// GaussMethodType; cov->key not given
{ // only for error reporting !
model *sub = cov;
while(isAnyDollar(sub)) sub = sub->sub[0];
SPRINTF(KT->error_loc, "Searching a simulation method for '%.50s'",
NICK(sub));
}
location_rules(cov, locpref); // hier cov
mixed_rules(cov, locpref, pref, order);
if (PL >= PL_RECURSIVE) {
defn *N = DefList + nr;
LPRINT("\n%.50s:%.50s\n", NAME(cov), NAME(next));
for (i=0; i < Nothing; i++) {
LPRINT("%-10s: base=%1d/%1d locpref=%3d totpref=%5d\n",
METHOD_NAMES[i], N->pref[i], next->pref[i],
(2 * locpref[i] < LOC_PREF_NONE) ? LOC_PREF_NONE + 1 : locpref[i],
pref[i]);
}
LPRINT("initstandard %.50s (%d) [pref value %d]\n",
DefList[nr].name, nr, pref[order[Nothing -1]]);
}
all_PREF_NONE = true;
for (i=0; i<Nothing; i++) all_PREF_NONE &= pref[i] == PREF_NONE;
if (cov->key != NULL) COV_DELETE(&(cov->key), cov);
int err;
if ((err = covcpy(&(cov->key), next)) != NOERROR) RETURN_ERR(err);
if (all_PREF_NONE) {
err = pref[Nothing] == PREF_NONE ? ERRORINVALIDMODEL : ERRORODDMODEL;
goto ErrorHandling;
}
err = CERROROUTOFMETHODLIST; // in case none of the methods are available
for (i = Nothing - 1; i>=0 && pref[order[i]] > PREF_NONE; i--) {
// assert(i ==Nothing-1);
if (PL >= PL_BRANCHING && i < Nothing - 1) {
LPRINT("'%.50s%.50s' failed for '%.50s'\n",
CAT_TYPE_NAMES[GaussMethodType], METHOD_NAMES[unimeth],NICK(next));
}
zaehler++;
unimeth = (Methods) order[i];
if (PL >= PL_BRANCHING) {
LPRINT("trying '%.50s%.50s' for '%.50s'\n",
CAT_TYPE_NAMES[GaussMethodType], METHOD_NAMES[unimeth],NICK(next));
}
meth = gaussmethod[unimeth];
addModel(&(cov->key), meth);
SET_CALLING(cov->key, cov);
model *key = cov->key;
key->prevloc = PLoc(cov);
Types frame =
MODELNR(key) == AVERAGE_INTERN ? PoissonGaussType :
hasAnyEvaluationFrame(cov) ? cov->frame : GaussMethodType;
if ((err = CHECK(key, logicaldim, xdim, GaussMethodType, OWNDOM(0),
OWNISO(0), cov->vdim, frame)) == NOERROR) {
//PMI(key);
if ((err = STRUCT(key, NULL)) <= NOERROR) {
//PMI(cov);
// assert(SYSMODEL(SYSOF(key)) != DIRECT); printf("xx\n");
if (!key->initialised) {
if ((err = CHECK(key, logicaldim, xdim, GaussMethodType, OWNDOM(0),
OWNISO(0), cov->vdim, frame)) == NOERROR){
CHECKED;
key->method = unimeth;
NEW_STORAGE(gen);
err = INIT(key, isPoissonGauss(frame) ? 2 : 0, cov->Sgen);
if (err == NOERROR) {
if (PL >= PL_RECURSIVE) {
LPRINT("returning to higher level...\n");
}
goto Initialised;
}
}
}
}
if (PL >= PL_RECURSIVE) {
LPRINT("back from %.50s: err =%d [%d, %.50s]\n",
METHOD_NAMES[unimeth], err, unimeth, METHOD_NAMES[unimeth]);
if (PL > PL_ERRORS) {
char info[LENERRMSG]; errorMSG(err, info);
PRINTF("error in gauss (init): %.50s\n",info);
// BUG;//
}
} // else if (PL >= PL_BRANCHING) M ERR(err);
} else {
if (PL > PL_ERRORS) {
char info[LENERRMSG]; errorMSG(err,info);
PRINTF("error in check: %.50s\n", info);
}
}
key = key->sub[0];
COV_DELETE_WITHOUTSUB(&(cov->key), cov);
cov->key=key;
SET_CALLING(key, cov);
}
if (PL >= PL_RECURSIVE) {
char *PREF_FAILURE = KEYtypeOf(cov)->PREF_FAILURE;
STRCPY(PREF_FAILURE, "");
int p, lp;
#define NMAX 14
char lpd[255], pd[255], names[NMAX];
names[NMAX-1] = '\0';
for (i=0; i<Nothing; i++) {
lp = locpref[i] - LOC_PREF_NONE;
p = pref[i] - LOC_PREF_NONE;
if (lp>0) {
STRCPY(lpd, "");
} else {
SPRINTF(lpd, "%.50s (locp) ", FailureMsg[MAX(0, MIN(MAXFAILMSG-1, 1-lp))]);
}
if (p>0 || p == lp) {
if (lp>0) STRCPY(pd, "(method specific failure)");
else STRCPY(pd, ""); // STRCPY(pd, "(method specific failure)");
} else {
SPRINTF(pd, "%.50s (pref)", FailureMsg[MAX(0, MIN(MAXFAILMSG-1, 1-p))]);
}
strcopyN(names, METHOD_NAMES[i], NMAX-1);
SPRINTF(dummy, "%.*s %-13.13s: %.5s%.70s\n", 100 * (int) Nothing -100,
PREF_FAILURE, names, lpd, pd);
STRCPY(PREF_FAILURE, dummy);
}
}
if (err != NOERROR) goto ErrorHandling;
Initialised:
// nachfolgende Zeile muss stehen, da Init in struct aufgerufen!
err = gauss_init_settings(cov);
cov->simu.active = cov->initialised = err == NOERROR;
ErrorHandling:
if (err <= NOERROR || zaehler==0){
if (err <= NOERROR) {
if (cov->key == NULL) next->simu.active = true;
else cov->key->simu.active = true;
}
RETURN_ERR(err);
}
if (zaehler==1) {
SPRINTF(KT->error_loc, "Only 1 method found for '%.50s', namely '%.50s', that comes into question.\nHowever, this method failed (err=%d)", NICK(next), METHOD_NAMES[unimeth], err);
RETURN_ERR(err);
}
{ // only for error reporting !
model *sub = next;
while(isAnyDollar(sub)) sub = sub->sub[0];
SPRINTF(KT->error_loc, "searching a simulation method for '%.50s'",
NICK(sub));
}
RETURN_ERR(ERROROUTOFMETHODLIST);
}
int init_gaussprocess(model *cov, gen_storage *s) {
/// ACHTUNG struct_gaussprocess hat bereits init aufgerufen!
// direkter Aufruf, u.a. durch RPtbm, average, cutoff, intrinsic, hyperplane,
// nugget, spectral !!!
if (hasAnyEvaluationFrame(cov)) {
assert(cov->key == NULL);
if (isnowVariogram(cov->sub[0])) return NOERROR;
return INIT(cov->sub[0], 0, s);
}
model
*key = cov->key;
int err;
assert(key != NULL);
if ((err = INIT(key, 0, s)) != NOERROR) RETURN_ERR(err);
if ((err = gauss_init_settings(cov)) != NOERROR) {
RETURN_ERR(err);
}
cov->simu.active = key->simu.active = true;
RETURN_NOERROR;
}
void do_gaussprocess(model *cov, gen_storage *s) {
assert(s != NULL);
// reopened by internal_dogauss
errorloc_type errorloc_save;
double *res = cov->rf;
int i,
vdimtot = Gettotalpoints(cov) * VDIM0 ;
model *key = cov->key ;
KEY_type *KT = cov->base;
SAVE_GAUSS_TRAFO;
STRCPY( errorloc_save, KT->error_loc);
if (cov->simu.pair) {
for (i=0; i<vdimtot; i++) res[i] = -res[i];
cov->simu.pair = false;
return;
} else {
cov->simu.pair = GLOBAL.gauss.paired;
}
assert(key != NULL);
// falls gaussprocess als solcher aufgerufen, dann cov->Sgen != NULL
// falls ueber *_USER, dann i.a.(?!) nicht gesetzt, sondern s:
DO(key, cov->Sgen == NULL ? s : cov->Sgen);
// (x^\lambda_1-1)/\lambda_1+\lambda_2
BOXCOX_INVERSE;
STRCPY( KT->error_loc, errorloc_save);
}
//////////////////////////////////////////////////////////////////////
// Binary Gauss
#define BINARY_THRESHOLD (GAUSSPROC_LAST + 1)
void kappa_binaryprocess(int i, model VARIABLE_IS_NOT_USED *cov,
int *nr, int *nc){
kappaGProc(i, cov, nr, nc);
if (i == BINARY_THRESHOLD) *nr = SIZE_NOT_DETERMINED;
}
int checkbinaryprocess(model *cov) {
// Achtung! binary hat Doppelbedeutung: binary-Gaussian als default
// und binary angewandt auf x-bel. process.
model
*key = cov->key,
*next = cov->sub[0],
*sub = key != NULL ? key : next;
double v;
int err = NOERROR;
if (PisNULL(BINARY_THRESHOLD)) kdefault(cov, BINARY_THRESHOLD, 0);
if (key == NULL && !isProcess(next)) {
if ((err = checkgaussprocess(cov)) != NOERROR) {
RETURN_ERR(err);
}
COV(ZERO(sub), sub, &v);
if (v != 1.0)
SERR("binaryian requires a correlation function as submodel.");
} else {
if ((err = CHECK_PASSTF(sub, ProcessType, SUBMODEL_DEP,
hasAnyEvaluationFrame(cov) ? cov->frame :
NormedProcessType
)) != NOERROR)
RETURN_ERR(err);
// if ((err = CHECK(sub, cov->tsdim, cov->xdimprev, ProcessType,
// OWNDOM(0), OWNISO(0),
// SUBMODEL_DEP,
// frame)) != NOERROR) RETURN_ERR(err);
setbackward(cov, sub);
}
VDIM0 = sub->vdim[0];
VDIM1 = sub->vdim[1];
RETURN_NOERROR;
}
int struct_binaryprocess(model *cov, model VARIABLE_IS_NOT_USED **newmodel) {
model
*next = cov->sub[0];
int err;
if (isnowVariogram(next)) {
assert(cov->key == NULL);
err = covcpy(&(cov->key), cov);
// darauffolgende Zeile absichern:
if (DefList[COVNR].kappas != 3 || DefList[GAUSSPROC].kappas != 2) BUG;
if (cov->key != NULL && !PARAMisNULL(cov->key, BINARY_THRESHOLD)) {
PARAMFREE(cov->key, BINARY_THRESHOLD);
PARAMtoNULL(cov->key, BINARY_THRESHOLD);
}
if (err != NOERROR) RETURN_ERR(err); //!!cov
SET_NR(cov->key, GAUSSPROC);
err = CHECK_PASSTF(cov->key, ProcessType, SUBMODEL_DEP, GaussMethodType);
// err = CHECK(cov->key, cov->tsdim, cov->xdimprev, ProcessType,
// OWNDOM(0), OWNISO(0), SUBMODEL_DEP, GaussMethodType);
if (err != NOERROR) RETURN_ERR(err);
err = STRUCT(cov->key, NULL);
RETURN_ERR(err);
}
else return STRUCT(next, NULL);
}
int init_binaryprocess( model *cov, gen_storage *s) {
double sigma,
*mean = NULL,
*variance = NULL,
*p = P(BINARY_THRESHOLD);
model
*key = cov->key,
*next=cov->sub[0],
*sub = key != NULL ? key : next;
int v, w, nmP1, pi,
npi = cov->nrow[BINARY_THRESHOLD],
err = NOERROR,
vdim = next->vdim[0],
vdimSq = vdim * vdim,
vdimP1 = vdim + 1;
assert(next->vdim[0] == next->vdim[1]);
if ((variance = (double*) MALLOC(sizeof(double) * vdimSq)) == NULL ||
(mean = (double*) CALLOC(vdim, sizeof(double))) == NULL
) goto ErrorHandling;
if ((err = INIT(sub, 0, s)) != NOERROR) goto ErrorHandling;
cov->rf = sub->rf;
cov->origrf = false;
if (isnowVariogram(next) || NEXTNR == GAUSSPROC) {
GetInternalMean(next, vdim, mean);
if (ISNAN(mean[0])) // GetInternalMean currently only allows ...
GERR1("'%.50s' currently only allows scalar fields - NA returned",
NICK(cov));
if (cov->mpp.moments >= 1) {
model *sub0 = NEXTNR==GAUSSPROC ? next->sub[0]: next;
COV(ZERO(sub0), sub0, variance);
}
nmP1 = cov->mpp.moments + 1;
for (pi=v=w=0; w<vdimSq; w+=vdimP1, v++, pi = (pi + 1) % npi ) {
int idx = v * nmP1;
cov->mpp.maxheights[v] = 1.0;
if (cov->mpp.moments >= 0) {
assert(cov->mpp.mM != NULL);
cov->mpp.mM[idx + 0] = cov->mpp.mMplus[idx + 0] = 1.0;
if (cov->mpp.moments >= 1) {
if (variance[w]==0.0)
GERR1("Vanishing sill not allowed in '%.50s'", NICK(next));
sigma = SQRT(variance[w]);
cov->mpp.mM[idx + 1] = cov->mpp.mMplus[idx + 1] =
pnorm(p[pi], mean[v], sigma, false, false);
int i;
for (i=2; i<= cov->mpp.moments; i++)
cov->mpp.mM[idx + i] = cov->mpp.mMplus[idx + i] = cov->mpp.mM[idx + 1];
}
}
}
}
cov->fieldreturn = wahr;
cov->simu.active = err == NOERROR;
ErrorHandling:
FREE(variance);
FREE(mean);
RETURN_ERR(err);
}
void do_binaryprocess(model *cov, gen_storage *s){
// reopened by internal_dogauss
long j,
tot = Loc(cov)->totalpoints,
endfor = tot;
assert(cov->ownloc == NULL);
int i,
pi,
npi = cov->nrow[BINARY_THRESHOLD],
vdim = VDIM0;
double
threshold,
*p = P(BINARY_THRESHOLD),
*rf = cov->rf;
model
*next=cov->sub[0];
assert(VDIM0 == VDIM1);
if (isnowVariogram(next)) {
do_gaussprocess(cov, s);
} else DO(next, s);
for (j=pi=i=0; i<vdim; i++, pi = (pi + 1) % npi, endfor += tot) {
threshold = p[pi];
// printf("endfor %d %10g\n", endfor, threshold);
if (threshold > RF_NEGINF && threshold < RF_INF)
for ( ; j<endfor; j++) rf[j] = (double) (rf[j] >= threshold);
}
}
void rangebinaryprocess(model *cov, range_type *range) {
rangegaussprocess(cov, range);
range->min[BINARY_THRESHOLD] = RF_NEGINF;
range->max[BINARY_THRESHOLD] = RF_INF;
range->pmin[BINARY_THRESHOLD] = -3;
range->pmax[BINARY_THRESHOLD] = 3;
range->openmin[BINARY_THRESHOLD] = false;
range->openmax[BINARY_THRESHOLD] = false;
}
//////////////////////////////////////////////////////////////////////
// Chisq Gauss
#define CHISQ_DEGREE 1
#define T_NU CHISQ_DEGREE
int checkchisqprocess(model *cov) {
ASSERT_ONESYSTEM;
model
*key = cov->key,
*next = cov->sub[0];
double *v = NULL;
int err = NOERROR,
xdim = OWNXDIM(0), // could differ from logicaldim in case of distances!
logicaldim = OWNLOGDIM(0);
assert((Loc(cov)->distances && xdim==1) || xdim == logicaldim);
if (PisNULL(CHISQ_DEGREE)) SERR("degree of freedom must be given");
if (key == NULL) {
// if (!isGaussMethod(next) && !isVariogram(next))
// SERR1("Gaussian process required, but '%.50s' obtained", NICK(cov));
if ((err = CHECK(next, logicaldim, xdim, ProcessType, XONLY, OWNISO(0),
SUBMODEL_DEP, GaussMethodType)) != NOERROR) {
if ((err = CHECK(next, logicaldim, xdim,
isCartesian(OWNISO(0)) ? VariogramType : PosDefType,
KERNEL,
CoordinateSystemOf(OWNISO(0)),
SUBMODEL_DEP, GaussMethodType)) != NOERROR) {
RETURN_ERR(err);
}
}
int
vdim = next->vdim[0],
vdimSq = vdim * vdim;
assert(vdim > 0);
if ((v = (double*) MALLOC(sizeof(double) * vdimSq)) == NULL)
RETURN_ERR(ERRORMEMORYALLOCATION);
model *sub = next;
while (sub != NULL && isnowProcess(sub)) sub=sub->sub[0];
if (sub == NULL) BUG;
if (isXonly(PREVSYSOF(sub))) {
COV(ZERO(next), next, v);
} else {
assert(isKernel(PREVSYSOF(sub)));
double *zero = ZERO(next);
NONSTATCOV(zero, zero, next, v);
}
int w,
vdimP1 = vdim + 1;
for (w=0; w<vdimSq; w+=vdimP1) {
if (v[w] != 1.0) {
FREE(v);
SERR("chisq requires a correlation function as submodel.");
}
}
FREE(v);
VDIM0 = sub->vdim[0];
VDIM1 = sub->vdim[1];
//no need to setbackward
} else {
if ((err = CHECK_PASSTF(key, ProcessType, SUBMODEL_DEP, GaussMethodType))
!= NOERROR) RETURN_ERR(err);
// if ((err = CHECK(key, cov->tsdim, cov->xdimprev, ProcessType,
// OWNDOM(0), OWNISO(0),
// SUBMODEL_DEP, cov->frame)) != NOERROR) RETURN_ERR(err);
setbackward(cov, key);
}
if ((err = kappaBoxCoxParam(cov, GAUSS_BOXCOX)) != NOERROR) RETURN_ERR(err);
RETURN_NOERROR;
}
int struct_chisqprocess(model *cov,
model VARIABLE_IS_NOT_USED **newmodel) {
model
*next = cov->sub[0];
int err;
if (isnowVariogram(next)) {
assert(cov->key == NULL);
if ((err = covcpy(&(cov->key), next)) > NOERROR) RETURN_ERR(err);
addModel(&(cov->key), GAUSSPROC);
SET_CALLING(cov->key, cov);
if ((err = CHECK_PASSFRAME(cov->key, GaussMethodType)) != NOERROR) RETURN_ERR(err);
// if ((err = CHECK(cov->key, cov->tsdim, cov->tsdim, ProcessType,
// cov->domprev, cov->isoprev,
// SUBMODEL_DEP, GaussMethodType)) != NOERROR) RETURN_ERR(err);
return STRUCT(cov->key, NULL);
} else {
return STRUCT(next, NULL);
}
}
int init_chisqprocess(model *cov, gen_storage *s) {
double // sigma,
mean, m2,
variance;
model
*key = cov->key,
*next = cov->sub[0],
*sub = key == NULL ? next : key;
int i, err, nmP1,
nmP1sub = sub->mpp.moments + 1,
vdim = VDIM0;
cov->simu.active = false;
if ((err = INIT(sub,
DefList[COVNR].range == rangechisqprocess ? 2 :
DefList[COVNR].range == rangetprocess ? 1 : 9999
, s)) != NOERROR) RETURN_ERR(err);
nmP1 = cov->mpp.moments + 1;
for (i=0; i<vdim; i++) {
int
idx = i * nmP1,
idxsub = i * nmP1sub;
mean = sub->mpp.mM[idxsub + 1];
m2 = sub->mpp.mM[idxsub + 2];
variance = m2 - mean * mean;
if (variance==0.0)
SERR1("Vanishing sill not allowed in '%.50s'", NICK(sub));
if (ISNAN(mean)) // GetInternalMean currently only allows ...
SERR1("'%.50s' currently only allows scalar fields -- NA returned",
NICK(cov));
cov->mpp.maxheights[i] = GLOBAL.extreme.standardmax *
GLOBAL.extreme.standardmax * m2;
if (cov->mpp.moments >= 0) {
assert(cov->mpp.mM != NULL);
cov->mpp.mM[idx + 0] = cov->mpp.mMplus[idx + 0] = 1.0;
if (cov->mpp.moments >= 1) {
cov->mpp.mMplus[idx + 1] =
DefList[COVNR].range == rangechisqprocess ? m2
: DefList[COVNR].range == rangetprocess ? RF_NAN
: RF_NAN;
cov->mpp.mM[idx + 1] = RF_NA; // to do: richtiger Wert
if (cov->mpp.moments >= 2)
cov->mpp.mM[idx + 2] = 3 * variance * RF_NA;//check the 4th moment of a Gaussian, to do
}
}
}
if (DefList[COVNR].range == rangechisqprocess) ReturnOwnField(cov);
else if (DefList[COVNR].range == rangetprocess) ReturnOtherField(cov, sub);
else BUG;
cov->simu.active = true;
RETURN_NOERROR;
}
void do_chisqprocess(model *cov, gen_storage *s){
// reopened by internal_dogauss
int i, f,
degree = P0INT(CHISQ_DEGREE),
totvdim = Loc(cov)->totalpoints * VDIM0;
assert(cov->ownloc == NULL);
model
*next=cov->sub[0],
*key = cov->key,
*sub = key == NULL ? next : key;
double
*subrf = sub->rf,
*res = cov->rf;
for (i=0; i<totvdim; i++) res[i] = 0.0;
for (f=0; f<degree; f++) {
DO(sub, s);
for (i=0; i<totvdim; i++) {
double dummy = subrf[i];
res[i] += dummy * dummy;
}
}
BOXCOX_INVERSE;
}
void rangechisqprocess(model VARIABLE_IS_NOT_USED *cov, range_type *range) {
GAUSS_COMMON_RANGE;
range->min[CHISQ_DEGREE] = 1;
range->max[CHISQ_DEGREE] = RF_INF;
range->pmin[CHISQ_DEGREE] = 1;
range->pmax[CHISQ_DEGREE] = 1000;
range->openmin[CHISQ_DEGREE] = false;
range->openmax[CHISQ_DEGREE] = true;
}
//////////////////////////////////////////////////////////////////////
// multivariate t distribution
void do_tprocess(model *cov, gen_storage *s){
// reopened by internal_dogauss
int i,
n = Loc(cov)->totalpoints * VDIM0;
model
*next=cov->sub[0],
*key = cov->key,
*sub = key == NULL ? next : key;
double
nu = P0(T_NU),
factor = SQRT(nu / rgamma(0.5 * nu, 0.5)), // note inverse!! see def rgamma
*res = cov->rf;
DO(sub, s);
for (i=0; i<n; res[i++] *= factor);
BOXCOX_INVERSE;
}
void rangetprocess(model VARIABLE_IS_NOT_USED *cov, range_type *range) {
GAUSS_COMMON_RANGE;
range->min[T_NU] = 0;
range->max[T_NU] = RF_INF;
range->pmin[T_NU] = 1;
range->pmax[T_NU] = 1000;
range->openmin[T_NU] = true;
range->openmax[T_NU] = true;
}