--- imach/src/imach.c 2022/05/15 15:06:23 1.317 +++ imach/src/imach.c 2022/07/27 14:47:35 1.327 @@ -1,6 +1,43 @@ -/* $Id: imach.c,v 1.317 2022/05/15 15:06:23 brouard Exp $ +/* $Id: imach.c,v 1.327 2022/07/27 14:47:35 brouard Exp $ $State: Exp $ $Log: imach.c,v $ + Revision 1.327 2022/07/27 14:47:35 brouard + Summary: Still a problem for one-step probabilities in case of quantitative variables + + Revision 1.326 2022/07/26 17:33:55 brouard + Summary: some test with nres=1 + + Revision 1.325 2022/07/25 14:27:23 brouard + Summary: r30 + + * imach.c (Module): Error cptcovn instead of nsd in bmij (was + coredumped, revealed by Feiuno, thank you. + + Revision 1.324 2022/07/23 17:44:26 brouard + *** empty log message *** + + Revision 1.323 2022/07/22 12:30:08 brouard + * imach.c (Module): Output of Wald test in the htm file and not only in the log. + + Revision 1.322 2022/07/22 12:27:48 brouard + * imach.c (Module): Output of Wald test in the htm file and not only in the log. + + Revision 1.321 2022/07/22 12:04:24 brouard + Summary: r28 + + * imach.c (Module): Output of Wald test in the htm file and not only in the log. + + Revision 1.320 2022/06/02 05:10:11 brouard + *** empty log message *** + + Revision 1.319 2022/06/02 04:45:11 brouard + * imach.c (Module): Adding the Wald tests from the log to the main + htm for better display of the maximum likelihood estimators. + + Revision 1.318 2022/05/24 08:10:59 brouard + * imach.c (Module): Some attempts to find a bug of wrong estimates + of confidencce intervals with product in the equation modelC + Revision 1.317 2022/05/15 15:06:23 brouard * imach.c (Module): Some minor improvements @@ -847,7 +884,7 @@ The same imach parameter file can be used but the option for mle should be -3. - Agnès, who wrote this part of the code, tried to keep most of the + Agnès, who wrote this part of the code, tried to keep most of the former routines in order to include the new code within the former code. The output is very simple: only an estimate of the intercept and of @@ -1026,13 +1063,13 @@ Important routines - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities) and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually. - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables - o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if + o There are 2**cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless. - Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr). - Institut national d'études démographiques, Paris. + Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr). + Institut national d'études démographiques, Paris. This software have been partly granted by Euro-REVES, a concerted action from the European Union. It is copyrighted identically to a GNU software product, ie programme and @@ -1096,6 +1133,7 @@ Important routines #define POWELLNOF3INFF1TEST /* Skip test */ /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */ /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */ +/* #define FLATSUP *//* Suppresses directions where likelihood is flat */ #include #include @@ -1162,7 +1200,7 @@ typedef struct { #define NINTERVMAX 8 #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */ #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */ -#define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */ +#define NCOVMAX 30 /**< Maximum number of covariates used in the model, including generated covariates V1*V2 or V1*age */ #define codtabm(h,k) (1 & (h-1) >> (k-1))+1 /*#define decodtabm(h,k,cptcoveff)= (h <= (1<> (k-1)) & 1) +1 : -1)*/ #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1 @@ -1186,12 +1224,12 @@ typedef struct { #define ODIRSEPARATOR '\\' #endif -/* $Id: imach.c,v 1.317 2022/05/15 15:06:23 brouard Exp $ */ +/* $Id: imach.c,v 1.327 2022/07/27 14:47:35 brouard Exp $ */ /* $State: Exp $ */ #include "version.h" char version[]=__IMACH_VERSION__; -char copyright[]="May 2022,INED-EUROREVES-Institut de longevite-Japan Society for the Promotion of Science (Grant-in-Aid for Scientific Research 25293121), Intel Software 2015-2020, Nihon University 2021-202, INED 2000-2022"; -char fullversion[]="$Revision: 1.317 $ $Date: 2022/05/15 15:06:23 $"; +char copyright[]="July 2022,INED-EUROREVES-Institut de longevite-Japan Society for the Promotion of Science (Grant-in-Aid for Scientific Research 25293121), Intel Software 2015-2020, Nihon University 2021-202, INED 2000-2022"; +char fullversion[]="$Revision: 1.327 $ $Date: 2022/07/27 14:47:35 $"; char strstart[80]; char optionfilext[10], optionfilefiname[FILENAMELENGTH]; int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */ @@ -1372,19 +1410,48 @@ double ***cotvar; /* Time varying covari double ***cotqvar; /* Time varying quantitative covariate itqv */ double idx; int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */ -/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ -/*k 1 2 3 4 5 6 7 8 9 */ -/*Tvar[k]= 5 4 3 6 5 2 7 1 1 */ -/* Tndvar[k] 1 2 3 4 5 */ -/*TDvar 4 3 6 7 1 */ /* For outputs only; combination of dummies fixed or varying */ -/* Tns[k] 1 2 2 4 5 */ /* Number of single cova */ -/* TvarsD[k] 1 2 3 */ /* Number of single dummy cova */ -/* TvarsDind 2 3 9 */ /* position K of single dummy cova */ -/* TvarsQ[k] 1 2 */ /* Number of single quantitative cova */ -/* TvarsQind 1 6 */ /* position K of single quantitative cova */ -/* Tprod[i]=k 4 7 */ -/* Tage[i]=k 5 8 */ -/* */ +/* Some documentation */ + /* Design original data + * V1 V2 V3 V4 V5 V6 V7 V8 Weight ddb ddth d1st s1 V9 V10 V11 V12 s2 V9 V10 V11 V12 + * < ncovcol=6 > nqv=2 (V7 V8) dv dv dv qtv dv dv dvv qtv + * ntv=3 nqtv=1 + * cptcovn number of covariates (not including constant and age) = # of + plus 1 = 10+1=11 + * For time varying covariate, quanti or dummies + * cotqvar[wav][iv(1 to nqtv)][i]= [1][12][i]=(V12) quanti + * cotvar[wav][ntv+iv][i]= [3+(1 to nqtv)][i]=(V12) quanti + * cotvar[wav][iv(1 to ntv)][i]= [1][1][i]=(V9) dummies at wav 1 + * cotvar[wav][iv(1 to ntv)][i]= [1][2][i]=(V10) dummies at wav 1 + * covar[k,i], value of kth fixed covariate dummy or quanti : + * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8) + * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 + V9 + V9*age + V10 + * k= 1 2 3 4 5 6 7 8 9 10 11 + */ +/* According to the model, more columns can be added to covar by the product of covariates */ +/* ncovcol=1(Males=0 Females=1) nqv=1(raedyrs) ntv=2(withoutiadl=0 withiadl=1, witoutadl=0 withoutadl=1) nqtv=1(bmi) nlstate=3 ndeath=1 + # States 1=Coresidence, 2 Living alone, 3 Institution + # V1=sex, V2=raedyrs Quant Fixed, State=livarnb4..livarnb11, V3=iadl4..iald11, V4=adlw4..adlw11, V5=r4bmi..r11bmi +*/ +/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ +/* k 1 2 3 4 5 6 7 8 9 */ +/*Typevar[k]= 0 0 0 2 1 0 2 1 0 *//*0 for simple covariate (dummy, quantitative,*/ + /* fixed or varying), 1 for age product, 2 for*/ + /* product */ +/*Dummy[k]= 1 0 0 1 3 1 1 2 0 *//*Dummy[k] 0=dummy (0 1), 1 quantitative */ + /*(single or product without age), 2 dummy*/ + /* with age product, 3 quant with age product*/ +/*Tvar[k]= 5 4 3 6 5 2 7 1 1 */ +/* nsd 1 2 3 */ /* Counting single dummies covar fixed or tv */ +/*TvarsD[nsd] 4 3 1 */ /* ID of single dummy cova fixed or timevary*/ +/*TvarsDind[k] 2 3 9 */ /* position K of single dummy cova */ +/* nsq 1 2 */ /* Counting single quantit tv */ +/* TvarsQ[k] 5 2 */ /* Number of single quantitative cova */ +/* TvarsQind 1 6 */ /* position K of single quantitative cova */ +/* Tprod[i]=k 1 2 */ /* Position in model of the ith prod without age */ +/* cptcovage 1 2 */ /* Counting cov*age in the model equation */ +/* Tage[cptcovage]=k 5 8 */ /* Position in the model of ith cov*age */ +/* Tvard[1][1]@4={4,3,1,2} V4*V3 V1*V2 */ /* Position in model of the ith prod without age */ +/* TvarF TvarF[1]=Tvar[6]=2, TvarF[2]=Tvar[7]=7, TvarF[3]=Tvar[9]=1 ID of fixed covariates or product V2, V1*V2, V1 */ +/* TvarFind; TvarFind[1]=6, TvarFind[2]=7, TvarFind[3]=9 *//* Inverse V2(6) is first fixed (single or prod) */ /* Type */ /* V 1 2 3 4 5 */ /* F F V V V */ @@ -1395,17 +1462,21 @@ int *TvarsDind; int *TvarsQ; int *TvarsQind; -#define MAXRESULTLINES 10 +#define MAXRESULTLINESPONE 10+1 int nresult=0; int parameterline=0; /* # of the parameter (type) line */ -int TKresult[MAXRESULTLINES]; -int Tresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */ -int Tinvresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */ -int Tvresult[MAXRESULTLINES][NCOVMAX]; /* For dummy variable , variable # (output) */ -double Tqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */ -double Tqinvresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */ -int Tvqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , variable # (output) */ - +int TKresult[MAXRESULTLINESPONE]; +int Tresult[MAXRESULTLINESPONE][NCOVMAX];/* For dummy variable , value (output) */ +int Tinvresult[MAXRESULTLINESPONE][NCOVMAX];/* For dummy variable , value (output) */ +int Tvresult[MAXRESULTLINESPONE][NCOVMAX]; /* For dummy variable , variable # (output) */ +double Tqresult[MAXRESULTLINESPONE][NCOVMAX]; /* For quantitative variable , value (output) */ +double Tqinvresult[MAXRESULTLINESPONE][NCOVMAX]; /* For quantitative variable , value (output) */ +int Tvqresult[MAXRESULTLINESPONE][NCOVMAX]; /* For quantitative variable , variable # (output) */ + +/* ncovcol=1(Males=0 Females=1) nqv=1(raedyrs) ntv=2(withoutiadl=0 withiadl=1, witoutadl=0 withoutadl=1) nqtv=1(bmi) nlstate=3 ndeath=1 + # States 1=Coresidence, 2 Living alone, 3 Institution + # V1=sex, V2=raedyrs Quant Fixed, State=livarnb4..livarnb11, V3=iadl4..iald11, V4=adlw4..adlw11, V5=r4bmi..r11bmi +*/ /* int *TDvar; /\**< TDvar[1]=4, TDvarF[2]=3, TDvar[3]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 *\/ */ int *TvarF; /**< TvarF[1]=Tvar[6]=2, TvarF[2]=Tvar[7]=7, TvarF[3]=Tvar[9]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ int *TvarFind; /**< TvarFind[1]=6, TvarFind[2]=7, Tvarind[3]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ @@ -2351,12 +2422,6 @@ void powell(double p[], double **xi, int double fp,fptt; double *xits; int niterf, itmp; -#ifdef LINMINORIGINAL -#else - - flatdir=ivector(1,n); - for (j=1;j<=n;j++) flatdir[j]=0; -#endif pt=vector(1,n); ptt=vector(1,n); @@ -2366,16 +2431,16 @@ void powell(double p[], double **xi, int for (j=1;j<=n;j++) pt[j]=p[j]; rcurr_time = time(NULL); for (*iter=1;;++(*iter)) { - fp=(*fret); /* From former iteration or initial value */ ibig=0; del=0.0; rlast_time=rcurr_time; /* (void) gettimeofday(&curr_time,&tzp); */ rcurr_time = time(NULL); curr_time = *localtime(&rcurr_time); - printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout); - fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog); + printf("\nPowell iter=%d -2*LL=%.12f gain=%.12f=%.3g %ld sec. %ld sec.",*iter,*fret, fp-*fret,fp-*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout); + fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f gain=%.12f=%.3g %ld sec. %ld sec.",*iter,*fret, fp-*fret,fp-*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog); /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */ + fp=(*fret); /* From former iteration or initial value */ for (i=1;i<=n;i++) { fprintf(ficrespow," %.12lf", p[i]); } @@ -2480,14 +2545,14 @@ void powell(double p[], double **xi, int /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */ /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */ /* New value of last point Pn is not computed, P(n-1) */ - for(j=1;j<=n;j++) { - if(flatdir[j] >0){ - printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]); - fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]); - } - /* printf("\n"); */ - /* fprintf(ficlog,"\n"); */ + for(j=1;j<=n;j++) { + if(flatdir[j] >0){ + printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]); + fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]); } + /* printf("\n"); */ + /* fprintf(ficlog,"\n"); */ + } /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /\* Did we reach enough precision? *\/ */ if (2.0*fabs(fp-(*fret)) <= ftol) { /* Did we reach enough precision? */ /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */ @@ -2525,10 +2590,6 @@ void powell(double p[], double **xi, int } #endif -#ifdef LINMINORIGINAL -#else - free_ivector(flatdir,1,n); -#endif free_vector(xit,1,n); free_vector(xits,1,n); free_vector(ptt,1,n); @@ -2642,6 +2703,13 @@ void powell(double p[], double **xi, int } printf("\n"); fprintf(ficlog,"\n"); +#ifdef FLATSUP + free_vector(xit,1,n); + free_vector(xits,1,n); + free_vector(ptt,1,n); + free_vector(pt,1,n); + return; +#endif } #endif printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig); @@ -2726,23 +2794,28 @@ void powell(double p[], double **xi, int newm=savm; /* Covariates have to be included here again */ cov[2]=agefin; - if(nagesqr==1) - cov[3]= agefin*agefin;; + if(nagesqr==1){ + cov[3]= agefin*agefin; + } for (k=1; k<=nsd;k++) { /* For single dummy covariates only */ /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */ cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)]; + /* cov[++k1]=nbcode[TvarsD[k]][codtabm(ij,k)]; */ /* printf("prevalim Dummy combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */ } for (k=1; k<=nsq;k++) { /* For single varying covariates only */ /* Here comes the value of quantitative after renumbering k with single quantitative covariates */ - cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k]; + cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k]; + /* cov[++k1]=Tqresult[nres][k]; */ /* printf("prevalim Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */ } for (k=1; k<=cptcovage;k++){ /* For product with age */ - if(Dummy[Tvar[Tage[k]]]){ + if(Dummy[Tage[k]]==2){ /* dummy with age */ cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; - } else{ - cov[2+nagesqr+Tage[k]]=Tqresult[nres][k]; + /* cov[++k1]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */ + } else if(Dummy[Tage[k]]==3){ /* quantitative with age */ + cov[2+nagesqr+Tage[k]]=Tqresult[nres][k]; + /* cov[++k1]=Tqresult[nres][k]; */ } /* printf("prevalim Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */ } @@ -2751,14 +2824,18 @@ void powell(double p[], double **xi, int if(Dummy[Tvard[k][1]==0]){ if(Dummy[Tvard[k][2]==0]){ cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; + /* cov[++k1]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */ }else{ cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k]; + /* cov[++k1]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k]; */ } }else{ if(Dummy[Tvard[k][2]==0]){ cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]]; + /* cov[++k1]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]]; */ }else{ cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]]; + /* cov[++k1]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]]; */ } } } @@ -2767,7 +2844,7 @@ void powell(double p[], double **xi, int /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/ /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */ /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */ - /* age and covariate values of ij are in 'cov' */ + /* age and covariate values of ij are in 'cov' */ out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */ savm=oldm; @@ -2890,8 +2967,9 @@ void powell(double p[], double **xi, int /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */ /* Covariates have to be included here again */ cov[2]=agefin; - if(nagesqr==1) + if(nagesqr==1){ cov[3]= agefin*agefin;; + } for (k=1; k<=nsd;k++) { /* For single dummy covariates only */ /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */ cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)]; @@ -2912,10 +2990,11 @@ void powell(double p[], double **xi, int /* /\* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; *\/ */ /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */ for (k=1; k<=cptcovage;k++){ /* For product with age */ - if(Dummy[Tvar[Tage[k]]]){ + /* if(Dummy[Tvar[Tage[k]]]== 2){ /\* dummy with age *\/ ERROR ???*/ + if(Dummy[Tage[k]]== 2){ /* dummy with age */ cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; - } else{ - cov[2+nagesqr+Tage[k]]=Tqresult[nres][k]; + } else if(Dummy[Tage[k]]== 3){ /* quantitative with age */ + cov[2+nagesqr+Tage[k]]=Tqresult[nres][k]; } /* printf("prevalim Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */ } @@ -2985,7 +3064,7 @@ void powell(double p[], double **xi, int maxmax=0.; for(i=1; i<=nlstate; i++){ - meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */ + meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column, could be nan! */ maxmax=FMAX(maxmax,meandiff[i]); /* printf("Back age= %d meandiff[%d]=%f, agefin=%d max[%d]=%f min[%d]=%f maxmax=%f\n", (int)age, i, meandiff[i],(int)agefin, i, max[i], i, min[i],maxmax); */ } /* i loop */ @@ -3071,7 +3150,7 @@ double **pmij(double **ps, double *cov, ps[i][i]=1./(s1+1.); /* Computing other pijs */ for(j=1; j> (k-1))+1 */ +/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ +/* k 1 2 3 4 5 6 7 8 9 */ +/*Tvar[k]= 5 4 3 6 5 2 7 1 1 */ +/* nsd 1 2 3 */ /* Counting single dummies covar fixed or tv */ +/*TvarsD[nsd] 4 3 1 */ /* ID of single dummy cova fixed or timevary*/ +/*TvarsDind[k] 2 3 9 */ /* position K of single dummy cova */ cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)]; /* printf("hpxij Dummy combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */ } for (k=1; k<=nsq;k++) { /* For single varying covariates only */ /* Here comes the value of quantitative after renumbering k with single quantitative covariates */ - cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k]; + cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k]; /* printf("hPxij Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */ } - for (k=1; k<=cptcovage;k++){ - if(Dummy[Tvar[Tage[k]]]){ + for (k=1; k<=cptcovage;k++){ /* For product with age V1+V1*age +V4 +age*V3 */ + /* 1+2 Tage[1]=2 TVar[2]=1 Dummy[2]=2, Tage[2]=4 TVar[4]=3 Dummy[4]=3 quant*/ + /* */ + if(Dummy[Tage[k]]== 2){ /* dummy with age */ + /* if(Dummy[Tvar[Tage[k]]]== 2){ /\* dummy with age *\/ */ cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; - } else{ - cov[2+nagesqr+Tage[k]]=Tqresult[nres][k]; + } else if(Dummy[Tage[k]]== 3){ /* quantitative with age */ + cov[2+nagesqr+Tage[k]]=Tqresult[nres][k]; } /* printf("hPxij Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */ } - for (k=1; k<=cptcovprod;k++){ /* */ + for (k=1; k<=cptcovprod;k++){ /* For product without age */ /* printf("hPxij Prod ij=%d k=%d Tprod[%d]=%d Tvard[%d][1]=V%d, Tvard[%d][2]=V%d\n",ij,k,k,Tprod[k], k,Tvard[k][1], k,Tvard[k][2]); */ - cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; + /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */ + if(Dummy[Tvard[k][1]==0]){ + if(Dummy[Tvard[k][2]==0]){ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; + }else{ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k]; + } + }else{ + if(Dummy[Tvard[k][2]==0]){ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]]; + }else{ + cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]]; + } + } } /* for (k=1; k<=cptcovn;k++) */ /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */ @@ -3373,7 +3478,7 @@ double ***hpxij(double ***po, int nhstep /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/ /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/ - /* right multiplication of oldm by the current matrix */ + /* right multiplication of oldm by the current matrix */ out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, pmij(pmmij,cov,ncovmodel,x,nlstate)); /* if((int)age == 70){ */ @@ -3443,13 +3548,15 @@ double ***hbxij(double ***po, int nhstep cov[1]=1.; agexact=age-( (h-1)*hstepm + (d) )*stepm/YEARM; /* age just before transition, d or d-1? */ /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */ + /* Debug */ + /* printf("hBxij age=%lf, agexact=%lf\n", age, agexact); */ cov[2]=agexact; if(nagesqr==1) cov[3]= agexact*agexact; - for (k=1; k<=cptcovn;k++){ + for (k=1; k<=nsd;k++){ /* For single dummy covariates only *//* cptcovn error */ /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */ /* /\* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; *\/ */ - cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)]; + cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];/* Bug valgrind */ /* printf("hbxij Dummy agexact=%.0f combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov[%d]=%lf codtabm(%d,Tvar[%d])=%d \n",agexact,ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],2+nagesqr+TvarsDind[k],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */ } for (k=1; k<=nsq;k++) { /* For single varying covariates only */ @@ -3457,16 +3564,30 @@ double ***hbxij(double ***po, int nhstep cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k]; /* printf("hPxij Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */ } - for (k=1; k<=cptcovage;k++){ /* Should start at cptcovn+1 */ - if(Dummy[Tvar[Tage[k]]]){ + for (k=1; k<=cptcovage;k++){ /* Should start at cptcovn+1 *//* For product with age */ + /* if(Dummy[Tvar[Tage[k]]]== 2){ /\* dummy with age error!!!*\/ */ + if(Dummy[Tage[k]]== 2){ /* dummy with age */ cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; - } else{ + } else if(Dummy[Tage[k]]== 3){ /* quantitative with age */ cov[2+nagesqr+Tage[k]]=Tqresult[nres][k]; } /* printf("hBxij Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */ } for (k=1; k<=cptcovprod;k++){ /* Useless because included in cptcovn */ cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; + if(Dummy[Tvard[k][1]==0]){ + if(Dummy[Tvard[k][2]==0]){ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; + }else{ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k]; + } + }else{ + if(Dummy[Tvard[k][2]==0]){ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]]; + }else{ + cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]]; + } + } } /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/ /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/ @@ -3476,7 +3597,7 @@ double ***hbxij(double ***po, int nhstep /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */ /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */ out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\ - 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); + 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);/* Bug valgrind */ /* if((int)age == 70){ */ /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */ /* for(i=1; i<=nlstate+ndeath; i++) { */ @@ -3562,11 +3683,16 @@ double func( double *x) */ ioffset=2+nagesqr ; /* Fixed */ - for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */ - cov[ioffset+TvarFind[k]]=covar[Tvar[TvarFind[k]]][i];/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V1 is fixed (k=6)*/ + for (k=1; k<=ncovf;k++){ /* For each fixed covariate dummu or quant or prod */ + /* # V1=sex, V2=raedyrs Quant Fixed, State=livarnb4..livarnb11, V3=iadl4..iald11, V4=adlw4..adlw11, V5=r4bmi..r11bmi */ + /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ + /* TvarF[1]=Tvar[6]=2, TvarF[2]=Tvar[7]=7, TvarF[3]=Tvar[9]=1 ID of fixed covariates or product V2, V1*V2, V1 */ + /* TvarFind; TvarFind[1]=6, TvarFind[2]=7, TvarFind[3]=9 *//* Inverse V2(6) is first fixed (single or prod) */ + cov[ioffset+TvarFind[k]]=covar[Tvar[TvarFind[k]]][i];/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V1 is fixed (TvarFind[1]=6)*/ + /* V1*V2 (7) TvarFind[2]=7, TvarFind[3]=9 */ } /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4] - is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2] + is 5, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]=6 has been calculated etc */ /* For an individual i, wav[i] gives the number of effective waves */ /* We compute the contribution to Likelihood of each effective transition @@ -3578,8 +3704,8 @@ double func( double *x) meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i] */ for(mi=1; mi<= wav[i]-1; mi++){ - for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/ - /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */ + for(k=1; k <= ncovv ; k++){ /* Varying covariates in the model (single and product but no age )"V5+V4+V3+V4*V3+V5*age+V1*age+V1" +TvarVind 1,2,3,4(V4*V3) Tvar[1]@7{5, 4, 3, 6, 5, 1, 1 ; 6 because the created covar is after V5 and is 6, minus 1+1, 3,2,1,4 positions in cotvar*/ + /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; but where is the crossproduct? */ cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i]; } for (ii=1;ii<=nlstate+ndeath;ii++) @@ -3594,10 +3720,10 @@ double func( double *x) if(nagesqr==1) cov[3]= agexact*agexact; /* Should be changed here */ for (kk=1; kk<=cptcovage;kk++) { - if(!FixedV[Tvar[Tage[kk]]]) - cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */ - else - cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact; + if(!FixedV[Tvar[Tage[kk]]]) + cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */ + else + cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact; } out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); @@ -3705,8 +3831,13 @@ double func( double *x) } /* end of individual */ } else if(mle==2){ for (i=1,ipmx=0, sw=0.; i<=imx; i++){ - for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; + ioffset=2+nagesqr ; + for (k=1; k<=ncovf;k++) + cov[ioffset+TvarFind[k]]=covar[Tvar[TvarFind[k]]][i]; for(mi=1; mi<= wav[i]-1; mi++){ + for(k=1; k <= ncovv ; k++){ + cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i]; + } for (ii=1;ii<=nlstate+ndeath;ii++) for (j=1;j<=nlstate+ndeath;j++){ oldm[ii][j]=(ii==j ? 1.0 : 0.0); @@ -4063,7 +4194,7 @@ void likelione(FILE *ficres,double p[], void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double [])) { - int i,j, iter=0; + int i,j,k, jk, jkk=0, iter=0; double **xi; double fret; double fretone; /* Only one call to likelihood */ @@ -4097,8 +4228,65 @@ void mlikeli(FILE *ficres,double p[], in if(j!=i)fprintf(ficrespow," p%1d%1d",i,j); fprintf(ficrespow,"\n"); #ifdef POWELL +#ifdef LINMINORIGINAL +#else /* LINMINORIGINAL */ + + flatdir=ivector(1,npar); + for (j=1;j<=npar;j++) flatdir[j]=0; +#endif /*LINMINORIGINAL */ + +#ifdef FLATSUP + powell(p,xi,npar,ftol,&iter,&fret,flatdir,func); + /* reorganizing p by suppressing flat directions */ + for(i=1, jk=1; i <=nlstate; i++){ + for(k=1; k <=(nlstate+ndeath); k++){ + if (k != i) { + printf("%d%d flatdir[%d]=%d",i,k,jk, flatdir[jk]); + if(flatdir[jk]==1){ + printf(" To be skipped %d%d flatdir[%d]=%d ",i,k,jk, flatdir[jk]); + } + for(j=1; j <=ncovmodel; j++){ + printf("%12.7f ",p[jk]); + jk++; + } + printf("\n"); + } + } + } +/* skipping */ + /* for(i=1, jk=1, jkk=1;(flatdir[jk]==0)&& (i <=nlstate); i++){ */ + for(i=1, jk=1, jkk=1;i <=nlstate; i++){ + for(k=1; k <=(nlstate+ndeath); k++){ + if (k != i) { + printf("%d%d flatdir[%d]=%d",i,k,jk, flatdir[jk]); + if(flatdir[jk]==1){ + printf(" To be skipped %d%d flatdir[%d]=%d jk=%d p[%d] ",i,k,jk, flatdir[jk],jk, jk); + for(j=1; j <=ncovmodel; jk++,j++){ + printf(" p[%d]=%12.7f",jk, p[jk]); + /*q[jjk]=p[jk];*/ + } + }else{ + printf(" To be kept %d%d flatdir[%d]=%d jk=%d q[%d]=p[%d] ",i,k,jk, flatdir[jk],jk, jkk, jk); + for(j=1; j <=ncovmodel; jk++,jkk++,j++){ + printf(" p[%d]=%12.7f=q[%d]",jk, p[jk],jkk); + /*q[jjk]=p[jk];*/ + } + } + printf("\n"); + } + fflush(stdout); + } + } + powell(p,xi,npar,ftol,&iter,&fret,flatdir,func); +#else /* FLATSUP */ powell(p,xi,npar,ftol,&iter,&fret,func); -#endif +#endif /* FLATSUP */ + +#ifdef LINMINORIGINAL +#else + free_ivector(flatdir,1,npar); +#endif /* LINMINORIGINAL*/ +#endif /* POWELL */ #ifdef NLOPT #ifdef NEWUOA @@ -4126,6 +4314,14 @@ void mlikeli(FILE *ficres,double p[], in } nlopt_destroy(opt); #endif +#ifdef FLATSUP + /* npared = npar -flatd/ncovmodel; */ + /* xired= matrix(1,npared,1,npared); */ + /* paramred= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel); */ + /* powell(pred,xired,npared,ftol,&iter,&fret,flatdir,func); */ + /* free_matrix(xire,1,npared,1,npared); */ +#else /* FLATSUP */ +#endif /* FLATSUP */ free_matrix(xi,1,npar,1,npar); fclose(ficrespow); printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p)); @@ -4577,7 +4773,7 @@ void freqsummary(char fileres[], double Title=%s
Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s
\n",\ fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model); } - fprintf(ficresphtm,"Current page is file %s
\n\n

Frequencies and prevalence by age at begin of transition and dummy covariate value at beginning of transition

\n",fileresphtm, fileresphtm); + fprintf(ficresphtm,"Current page is file %s
\n\n

Frequencies (weight=%d) and prevalence by age at begin of transition and dummy covariate value at beginning of transition

\n",fileresphtm, fileresphtm, weightopt); strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm")); if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) { @@ -4587,11 +4783,11 @@ Title=%s
Datafile=%s Firstpass=%d La exit(70); } else{ fprintf(ficresphtmfr,"\nIMaCh PHTM_Frequency table %s\n %s
%s
\ -
\n \ +,
\n \ Title=%s
Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s
\n",\ fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model); } - fprintf(ficresphtmfr,"Current page is file %s
\n\n

Frequencies of all effective transitions of the model, by age at begin of transition, and covariate value at the begin of transition (if the covariate is a varying covariate)

Unknown status is -1
\n",fileresphtmfr, fileresphtmfr); + fprintf(ficresphtmfr,"Current page is file %s
\n\n

(weight=%d) frequencies of all effective transitions of the model, by age at begin of transition, and covariate value at the begin of transition (if the covariate is a varying covariate)

Unknown status is -1
\n",fileresphtmfr, fileresphtmfr,weightopt); y= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE); x= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE); @@ -4769,7 +4965,7 @@ Title=%s
Datafile=%s Firstpass=%d La /* } */ } /* end bool */ } /* end iind = 1 to imx */ - /* prop[s][age] is feeded for any initial and valid live state as well as + /* prop[s][age] is fed for any initial and valid live state as well as freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */ @@ -5964,7 +6160,9 @@ void concatwav(int wav[], int **dh, int varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf; } } - + /* if((int)age ==50){ */ + /* printf(" age=%d cij=%d nres=%d varhe[%d][%d]=%f ",(int)age, cij, nres, 1,2,varhe[1][2]); */ + /* } */ /* Computing expectancies */ hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij,nres); for(i=1; i<=nlstate;i++) @@ -6627,6 +6825,7 @@ void varprob(char optionfilefiname[], do int k2, l2, j1, z1; int k=0, l; int first=1, first1, first2; + int nres=0; /* New */ double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp; double **dnewm,**doldm; double *xp; @@ -6715,6 +6914,8 @@ To be simple, these graphs help to under if (cptcovn<1) {tj=1;ncodemax[1]=1;} j1=0; for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/ + for(nres=1;nres <=1; nres++){ /* For each resultline */ + /* for(nres=1;nres <=nresult; nres++){ /\* For each resultline *\/ */ if (cptcovn>0) { fprintf(ficresprob, "\n#********** Variable "); for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]); @@ -6729,7 +6930,8 @@ To be simple, these graphs help to under fprintf(fichtmcov, "\n
********** Variable "); - for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]); + /* for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]); */ + for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtmcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]); fprintf(fichtmcov, "**********\n
"); fprintf(ficresprobcor, "\n#********** Variable "); @@ -6749,8 +6951,11 @@ To be simple, these graphs help to under cov[2]=age; if(nagesqr==1) cov[3]= age*age; - for (k=1; k<=cptcovn;k++) { - cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)]; + /* for (k=1; k<=cptcovn;k++) { */ + /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)]; */ + for (k=1; k<=nsd;k++) { /* For single dummy covariates only */ + /* Here comes the value of the covariate 'j1' after renumbering k with single dummy covariates */ + cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(j1,k)]; /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4 * 1 1 1 1 1 * 2 2 1 1 1 @@ -6758,12 +6963,42 @@ To be simple, these graphs help to under */ /* nbcode[1][1]=0 nbcode[1][2]=1;*/ } - /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */ - for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; - for (k=1; k<=cptcovprod;k++) - cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; - - + /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1, Tage[1]=2 */ + /* ) p nbcode[Tvar[Tage[k]]][(1 & (ij-1) >> (k-1))+1] */ + /*for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */ + for (k=1; k<=cptcovage;k++){ /* For product with age */ + if(Dummy[Tage[k]]==2){ /* dummy with age */ + cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(j1,k)]*cov[2]; + /* cov[++k1]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */ + } else if(Dummy[Tage[k]]==3){ /* quantitative with age */ + printf("Internal IMaCh error, don't know which value for quantitative covariate with age, Tage[k]%d, k=%d, Tvar[Tage[k]]=V%d, age=%d\n",Tage[k],k ,Tvar[Tage[k]], (int)cov[2]); + exit(1); + /* cov[2+nagesqr+Tage[k]]=meanq[k]/idq[k]*cov[2];/\* Using the mean of quantitative variable Tvar[Tage[k]] /\* Tqresult[nres][k]; *\/ */ + /* cov[++k1]=Tqresult[nres][k]; */ + } + /* cov[2+Tage[k]+nagesqr]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */ + } + for (k=1; k<=cptcovprod;k++){/* For product without age */ + if(Dummy[Tvard[k][1]==0]){ + if(Dummy[Tvard[k][2]==0]){ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(j1,k)] * nbcode[Tvard[k][2]][codtabm(j1,k)]; + /* cov[++k1]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */ + }else{ /* Should we use the mean of the quantitative variables? */ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(j1,k)] * Tqresult[nres][k]; + /* cov[++k1]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k]; */ + } + }else{ + if(Dummy[Tvard[k][2]==0]){ + cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(j1,k)] * Tqinvresult[nres][Tvard[k][1]]; + /* cov[++k1]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]]; */ + }else{ + cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]]; + /* cov[++k1]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]]; */ + } + } + /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; */ + } +/* For each age and combination of dummy covariates we slightly move the parameters of delti in order to get the gradient*/ for(theta=1; theta <=npar; theta++){ for(i=1; i<=npar; i++) xp[i] = x[i] + (i==theta ?delti[theta]:(double)0); @@ -6948,6 +7183,7 @@ To be simple, these graphs help to under } /* k12 */ } /*l1 */ }/* k1 */ + } /* loop on nres */ } /* loop on combination of covariates j1 */ free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage); free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage); @@ -6970,12 +7206,12 @@ void printinghtml(char fileresu[], char double jprev1, double mprev1,double anprev1, double dateprev1, double dateprojd, double dateback1, \ double jprev2, double mprev2,double anprev2, double dateprev2, double dateprojf, double dateback2){ int jj1, k1, i1, cpt, k4, nres; - + /* In fact some results are already printed in fichtm which is open */ fprintf(fichtm,""); - fprintf(fichtm,"
  • model=1+age+%s\n \ -
", model); +/* fprintf(fichtm,"
  • model=1+age+%s\n \ */ +/*
", model); */ fprintf(fichtm,"
  • Result files (first order: no variance)

    \n"); fprintf(fichtm,"
  • - Observed frequency between two states (during the period defined between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf): %s (html file)
    \n", jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm")); @@ -7077,7 +7313,7 @@ void printinghtml(char fileresu[], char } /* if(nqfveff+nqtveff 0) */ /* Test to be done */ - fprintf(fichtm," ************\n
    "); + fprintf(fichtm," (model=%s) ************\n
    ",model); if(invalidvarcomb[k1]){ fprintf(fichtm,"\n

    Combination (%d) ignored because no cases

    \n",k1); printf("\nCombination (%d) ignored because no cases \n",k1); @@ -7264,7 +7500,7 @@ See page 'Matrix of variance-covariance fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]); } - fprintf(fichtm," ************\n
    "); + fprintf(fichtm," (model=%s) ************\n
    ",model); if(invalidvarcomb[k1]){ fprintf(fichtm,"\n

    Combination (%d) ignored because no cases

    \n",k1); @@ -7400,7 +7636,7 @@ void printinggnuplot(char fileresu[], ch fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1,nres); fprintf(ficgp,"\n#set out \"V_%s_%d-%d-%d.svg\" \n",optionfilefiname,cpt,k1,nres); /* fprintf(ficgp,"set label \"Alive state %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",cpt,gplotlabel); */ - fprintf(ficgp,"set title \"Alive state %d %s\" font \"Helvetica,12\"\n",cpt,gplotlabel); + fprintf(ficgp,"set title \"Alive state %d %s model=%s\" font \"Helvetica,12\"\n",cpt,gplotlabel,model); fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \nset ter svg size 640, 480\nplot [%.f:%.f] \"%s\" every :::%d::%d u 1:($2==%d ? $3:1/0) \"%%lf %%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),nres-1,nres-1,nres); /* fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \nset ter svg size 640, 480\nplot [%.f:%.f] \"%s\" every :::%d::%d u 1:($2==%d ? $3:1/0) \"%%lf %%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),k1-1,k1-1,nres); */ /* k1-1 error should be nres-1*/ @@ -7821,8 +8057,8 @@ set ter svg size 640, 480\nunset log y\n fprintf(ficgp,", '' "); /* l=(nlstate+ndeath)*(i-1)+1; */ l=(nlstate+ndeath)*(cpt-1)+1; /* fixed for i; cpt=1 1, cpt=2 1+ nlstate+ndeath, 1+2*(nlstate+ndeath) */ - /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */ - /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */ + /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */ + /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */ fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+i-1); /* To be verified */ /* for (j=2; j<= nlstate ; j ++) */ /* fprintf(ficgp,"+$%d",k+l+j-1); */ @@ -8172,9 +8408,9 @@ set ter svg size 640, 480\nunset log y\n for(j=1; j <=cptcovt; j++) { /* For each covariate of the simplified model */ /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */ if(cptcovage >0){ /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, 2 V5 and V1 */ - if(j==Tage[ij]) { /* Product by age To be looked at!!*/ + if(j==Tage[ij]) { /* Product by age To be looked at!!*//* Bug valgrind */ if(ij <=cptcovage) { /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, 2 V5 and V1 */ - if(DummyV[j]==0){ + if(DummyV[j]==0){/* Bug valgrind */ fprintf(ficgp,"+p%d*%d*x",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]);; }else{ /* quantitative */ fprintf(ficgp,"+p%d*%f*x",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* Tqinvresult in decoderesult */ @@ -9504,23 +9740,23 @@ int readdata(char datafile[], int firsto } if(lval <-1 || lval >1){ printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \ - Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \ + Should be a value of %d(nth) covariate of wave %d (0 should be the value for the reference and 1\n \ for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \ For example, for multinomial values like 1, 2 and 3,\n \ build V1=0 V2=0 for the reference value (1),\n \ V1=1 V2=0 for (2) \n \ and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \ output of IMaCh is often meaningless.\n \ - Exiting.\n",lval,linei, i,line,j); + Exiting.\n",lval,linei, i,line,iv,j); fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \ - Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \ + Should be a value of %d(nth) covariate of wave %d (0 should be the value for the reference and 1\n \ for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \ For example, for multinomial values like 1, 2 and 3,\n \ build V1=0 V2=0 for the reference value (1),\n \ V1=1 V2=0 for (2) \n \ and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \ output of IMaCh is often meaningless.\n \ - Exiting.\n",lval,linei, i,line,j);fflush(ficlog); + Exiting.\n",lval,linei, i,line,iv,j);fflush(ficlog); return 1; } cotvar[j][iv][i]=(double)(lval); @@ -9748,33 +9984,32 @@ int decoderesult ( char resultline[], in return (0); } if( j != cptcovs ){ /* Be careful if a variable is in a product but not single */ - printf("ERROR: the number of variables in the resultline, %d, differs from the number of variables used in the model line, %d.\n",j, cptcovs); + printf("ERROR: the number of variables in this result line, %d, differs from the number of variables used in the model line, %d.\n",j, cptcovs); fprintf(ficlog,"ERROR: the number of variables in the resultline, %d, differs from the number of variables used in the model line, %d.\n",j, cptcovs); } for(k=1; k<=j;k++){ /* Loop on any covariate of the result line */ if(nbocc(resultsav,'=') >1){ - cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' ' - resultsav= V4=1 V5=25.1 V3=0 stra= V5=25.1 V3=0 strb= V4=1 */ - cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */ + cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' ' (stra is the rest of the resultline to be analyzed in the next loop *//* resultsav= "V4=1 V5=25.1 V3=0" stra= "V5=25.1 V3=0" strb= "V4=1" */ + cutl(strc,strd,strb,'='); /* strb:"V4=1" strc="1" strd="V4" */ }else cutl(strc,strd,resultsav,'='); - Tvalsel[k]=atof(strc); /* 1 */ + Tvalsel[k]=atof(strc); /* 1 */ /* Tvalsel of k is the float value of the kth covariate appearing in this result line */ cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */; - Tvarsel[k]=atoi(strc); + Tvarsel[k]=atoi(strc); /* 4 */ /* Tvarsel is the id of the kth covariate in the result line Tvarsel[1] in "V4=1.." is 4.*/ /* Typevarsel[k]=1; /\* 1 for age product *\/ */ /* cptcovsel++; */ if (nbocc(stra,'=') >0) strcpy(resultsav,stra); /* and analyzes it */ } /* Checking for missing or useless values in comparison of current model needs */ - for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ - if(Typevar[k1]==0){ /* Single covariate in model */ + for(k1=1; k1<= cptcovt ;k1++){ /* Loop on model. model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ + if(Typevar[k1]==0){ /* Single covariate in model *//*0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */ match=0; - for(k2=1; k2 <=j;k2++){/* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */ - if(Tvar[k1]==Tvarsel[k2]) {/* Tvar[1]=5 == Tvarsel[2]=5 */ + for(k2=1; k2 <=j;k2++){/* Loop on resultline. In result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */ + if(Tvar[k1]==Tvarsel[k2]) {/* Tvar is coming from the model, Tvarsel from the result. Tvar[1]=5 == Tvarsel[2]=5 */ modelresult[k2]=k1;/* modelresult[2]=1 modelresult[1]=2 modelresult[3]=3 modelresult[6]=4 modelresult[9]=5 */ - match=1; + match=1; /* modelresult of k2 variable of resultline is identical to k1 variable of the model good */ break; } } @@ -9786,12 +10021,12 @@ int decoderesult ( char resultline[], in } } /* Checking for missing or useless values in comparison of current model needs */ - for(k2=1; k2 <=j;k2++){ /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */ + for(k2=1; k2 <=j;k2++){ /* Loop on resultline variables: result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */ match=0; - for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ + for(k1=1; k1<= cptcovt ;k1++){ /* loop on model: model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ if(Typevar[k1]==0){ /* Single */ if(Tvar[k1]==Tvarsel[k2]) { /* Tvar[2]=4 == Tvarsel[1]=4 */ - resultmodel[k1]=k2; /* resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */ + resultmodel[k1]=k2; /* k2th variable of the model corresponds to k1 variable of the model. resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */ ++match; } } @@ -9825,7 +10060,7 @@ int decoderesult ( char resultline[], in /* V(Tvqresult)=Tqresult V5=25.1 V2=8 Tqresult[nres=1][1]=25.1 */ /* V5*age V5 known which value for nres? */ /* Tqinvresult[2]=8 Tqinvresult[1]=25.1 */ - for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* model line */ + for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* loop on model line */ if( Dummy[k1]==0 && Typevar[k1]==0 ){ /* Single dummy */ k3= resultmodel[k1]; /* resultmodel[2(V4)] = 1=k3 */ k2=(int)Tvarsel[k3]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */ @@ -9836,8 +10071,8 @@ int decoderesult ( char resultline[], in printf("Decoderesult Dummy k=%d, V(k2=V%d)= Tvalsel[%d]=%d, 2**(%d)\n",k, k2, k3, (int)Tvalsel[k3], k4); k4++;; } else if( Dummy[k1]==1 && Typevar[k1]==0 ){ /* Single quantitative */ - k3q= resultmodel[k1]; /* resultmodel[2] = 1=k3 */ - k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */ + k3q= resultmodel[k1]; /* resultmodel[1(V5)] = 25.1=k3q */ + k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[1]]= Tvarsel[1] = 4=k2 */ Tqresult[nres][k4q+1]=Tvalsel[k3q]; /* Tqresult[nres][1]=25.1 */ Tvqresult[nres][k4q+1]=(int)Tvarsel[k3q]; /* Tvqresult[nres][1]=5 */ Tqinvresult[nres][(int)Tvarsel[k3q]]=Tvalsel[k3q]; /* Tqinvresult[nres][5]=25.1 */ @@ -9860,11 +10095,12 @@ int decodemodel( char model[], int lasto * - cptcovs number of simple covariates * - Tvar[k] is the id of the kth covariate Tvar[1]@12 {1, 2, 3, 8, 10, 11, 8, 3, 7, 8, 5, 6}, thus Tvar[5=V7*V8]=10 * which is a new column after the 9 (ncovcol) variables. - * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual + * - if k is a product Vn*Vm, covar[k][i] is filled with correct values for each individual * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6. * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 . */ +/* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1, Tage[1]=2 */ { int i, j, k, ks, v; int j1, k1, k2, k3, k4; @@ -9942,12 +10178,12 @@ int decodemodel( char model[], int lasto * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2 * k= 1 2 3 4 5 6 7 8 9 10 11 12 * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8 - * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6} + * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6} * p Tprod[1]@2={ 6, 5} *p Tvard[1][1]@4= {7, 8, 5, 6} * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8 * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2]; - *How to reorganize? + *How to reorganize? Tvars(orted) * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6} * {2, 1, 4, 8, 5, 6, 3, 7} @@ -9972,22 +10208,23 @@ int decodemodel( char model[], int lasto Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0; } cptcovage=0; - for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */ - cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+' - modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */ - if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */ + for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model line */ + cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+' cutl from left to right + modelsav==V2+V1+V5*age+V4+V3*age strb=V3*age stra=V2+V1V5*age+V4 */ /* "V5+V4+V3+V4*V3+V5*age+V1*age+V1" strb="V5" stra="V4+V3+V4*V3+V5*age+V1*age+V1" */ + if (nbocc(modelsav,'+')==0) + strcpy(strb,modelsav); /* and analyzes it */ /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/ /*scanf("%d",i);*/ - if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */ - cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */ + if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V5*age+ V4+V3*age strb=V3*age */ + cutl(strc,strd,strb,'*'); /**< k=1 strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */ if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */ /* covar is not filled and then is empty */ cptcovprod--; cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */ - Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */ + Tvar[k]=atoi(stre); /* V2+V1+V5*age+V4+V3*age Tvar[5]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */ Typevar[k]=1; /* 1 for age product */ - cptcovage++; /* Sums the number of covariates which include age as a product */ - Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */ + cptcovage++; /* Counts the number of covariates which include age as a product */ + Tage[cptcovage]=k; /* V2+V1+V4+V3*age Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */ /*printf("stre=%s ", stre);*/ } else if (strcmp(strd,"age")==0) { /* or age*Vn */ cptcovprod--; @@ -10004,12 +10241,13 @@ int decodemodel( char model[], int lasto Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but because this model-covariate is a construction we invent a new column which is after existing variables ncovcol+nqv+ntv+nqtv + k1 - If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2 - Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */ + If already ncovcol=4 and model=V2 + V1 +V1*V4 +age*V3 +V3*V2 + thus after V4 we invent V5 and V6 because age*V3 will be computed in 4 + Tvar[3=V1*V4]=4+1=5 Tvar[5=V3*V2]=4 + 2= 6, Tvar[4=age*V3]=4 etc */ Typevar[k]=2; /* 2 for double fixed dummy covariates */ cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */ Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */ - Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */ + Tposprod[k]=k1; /* Tposprod[3]=1, Tposprod[2]=5 */ Tvard[k1][1] =atoi(strc); /* m 1 for V1*/ Tvard[k1][2] =atoi(stre); /* n 4 for V4*/ k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */ @@ -10024,7 +10262,7 @@ int decodemodel( char model[], int lasto } } /* End age is not in the model */ } /* End if model includes a product */ - else { /* no more sum */ + else { /* not a product */ /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/ /* scanf("%d",i);*/ cutl(strd,strc,strb,'V'); @@ -10055,7 +10293,7 @@ int decodemodel( char model[], int lasto model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place k = 1 2 3 4 5 6 7 8 9 Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5 - Typevar[k]= 0 0 0 2 1 0 2 1 1 + Typevar[k]= 0 0 0 2 1 0 2 1 0 Fixed[k] 1 1 1 1 3 0 0 or 2 2 3 Dummy[k] 1 0 0 0 3 1 1 2 3 Tmodelind[combination of covar]=k; @@ -10064,11 +10302,11 @@ int decodemodel( char model[], int lasto /* If Tvar[k] >ncovcol it is a product */ /* Tvar[k] is the value n of Vn with n varying for 1 to nvcol, or p Vp=Vn*Vm for product */ /* Computing effective variables, ie used by the model, that is from the cptcovt variables */ - printf("Model=%s\n\ + printf("Model=1+age+%s\n\ Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\ Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\ Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product\n",model); - fprintf(ficlog,"Model=%s\n\ + fprintf(ficlog,"Model=1+age+%s\n\ Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\ Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\ Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product\n",model); @@ -10135,7 +10373,7 @@ Dummy[k] 0=dummy (0 1), 1 quantitative ( modell[k].subtype= VQ; ncovv++; /* Only simple time varying variables */ nsq++; - TvarsQ[nsq]=Tvar[k]; + TvarsQ[nsq]=Tvar[k]; /* k=1 Tvar=5 nsq=1 TvarsQ[1]=5 */ TvarsQind[nsq]=k; TvarV[ncovv]=Tvar[k]; TvarVind[ncovv]=k; /* TvarVind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Any time varying singele */ @@ -11001,7 +11239,7 @@ int hPijx(double *p, int bage, int fage) /* oldm=oldms;savm=savms; */ /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */ - hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k, nres); + hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k, nres);/* Bug valgrind */ /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */ fprintf(ficrespijb,"# Cov Agex agex-h hbijx with i,j="); for(i=1; i<=nlstate;i++) @@ -11014,7 +11252,7 @@ int hPijx(double *p, int bage, int fage) /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */ for(i=1; i<=nlstate;i++) for(j=1; j<=nlstate+ndeath;j++) - fprintf(ficrespijb," %.5f", p3mat[i][j][h]); + fprintf(ficrespijb," %.5f", p3mat[i][j][h]);/* Bug valgrind */ fprintf(ficrespijb,"\n"); } free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); @@ -11067,6 +11305,7 @@ int main(int argc, char *argv[]) double dum=0.; /* Dummy variable */ double ***p3mat; /* double ***mobaverage; */ + double wald; char line[MAXLINE]; char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE]; @@ -11698,7 +11937,8 @@ Please run with mle=-1 to get a correct } mint=matrix(1,maxwav,firstobs,lastobs); anint=matrix(1,maxwav,firstobs,lastobs); - s=imatrix(1,maxwav+1,firstobs,lastobs); /* s[i][j] health state for wave i and individual j */ + s=imatrix(1,maxwav+1,firstobs,lastobs); /* s[i][j] health state for wave i and individual j */ + printf("BUG ncovmodel=%d NCOVMAX=%d 2**ncovmodel=%f BUG\n",ncovmodel,NCOVMAX,pow(2,ncovmodel)); tab=ivector(1,NCOVMAX); ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */ ncodemaxwundef=ivector(1,NCOVMAX); /* Number of code per covariate; if - 1 O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */ @@ -11975,7 +12215,7 @@ Title=%s
    Datafile=%s Firstpass=%d La optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model); } - fprintf(fichtm,"\n\n\nIMaCh %s\n IMaCh for Interpolated Markov Chain
    \nSponsored by Copyright (C) 2002-2015 INED-EUROREVES-Institut de longévité-2013-2016-Japan Society for the Promotion of Sciences 日本学術振興会 (Grant-in-Aid for Scientific Research 25293121) - Intel Software 2015-2018
    \ + fprintf(fichtm,"\n\n\nIMaCh %s\n IMaCh for Interpolated Markov Chain
    \nSponsored by Copyright (C) 2002-2015 INED-EUROREVES-Institut de longévité-2013-2016-Japan Society for the Promotion of Sciences 日本学術振興会 (Grant-in-Aid for Scientific Research 25293121) - Intel Software 2015-2018
    \
    \n\ IMaCh-%s
    %s
    \
    \n\ @@ -12311,48 +12551,130 @@ Please run with mle=-1 to get a correct fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n"); - printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n"); + printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n"); /* Printing model equation */ fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n"); + + printf("#model= 1 + age "); + fprintf(ficres,"#model= 1 + age "); + fprintf(ficlog,"#model= 1 + age "); + fprintf(fichtm,"\n
    • model=1+age+%s\n \ +
    ", model); + + fprintf(fichtm,"\n\n"); + fprintf(fichtm, ""); + if(nagesqr==1){ + printf(" + age*age "); + fprintf(ficres," + age*age "); + fprintf(ficlog," + age*age "); + fprintf(fichtm, ""); + } + for(j=1;j <=ncovmodel-2;j++){ + if(Typevar[j]==0) { + printf(" + V%d ",Tvar[j]); + fprintf(ficres," + V%d ",Tvar[j]); + fprintf(ficlog," + V%d ",Tvar[j]); + fprintf(fichtm, "",Tvar[j]); + }else if(Typevar[j]==1) { + printf(" + V%d*age ",Tvar[j]); + fprintf(ficres," + V%d*age ",Tvar[j]); + fprintf(ficlog," + V%d*age ",Tvar[j]); + fprintf(fichtm, "",Tvar[j]); + }else if(Typevar[j]==2) { + printf(" + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]); + fprintf(ficres," + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]); + fprintf(ficlog," + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]); + fprintf(fichtm, "",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]); + } + } + printf("\n"); + fprintf(ficres,"\n"); + fprintf(ficlog,"\n"); + fprintf(fichtm, ""); + fprintf(fichtm, "\n"); + + for(i=1,jk=1; i <=nlstate; i++){ for(k=1; k <=(nlstate+ndeath); k++){ if (k != i) { + fprintf(fichtm, ""); printf("%d%d ",i,k); fprintf(ficlog,"%d%d ",i,k); fprintf(ficres,"%1d%1d ",i,k); + fprintf(fichtm, "",i,k); for(j=1; j <=ncovmodel; j++){ printf("%12.7f ",p[jk]); fprintf(ficlog,"%12.7f ",p[jk]); fprintf(ficres,"%12.7f ",p[jk]); + fprintf(fichtm, "",p[jk]); jk++; } printf("\n"); fprintf(ficlog,"\n"); fprintf(ficres,"\n"); + fprintf(fichtm, "\n"); } } } + /* fprintf(fichtm,"\n"); */ + fprintf(fichtm,"
    Model=1+ age+ age*age+ V%d+ V%d*age+ V%d*V%d
    %1d%1d%12.7f
    \n"); + fprintf(fichtm, "\n"); + if(mle != 0){ /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */ ftolhess=ftol; /* Usually correct */ hesscov(matcov, hess, p, npar, delti, ftolhess, func); printf("Parameters and 95%% confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W .\n But be careful that parameters are highly correlated because incidence of disability is highly correlated to incidence of recovery.\n It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n"); fprintf(ficlog, "Parameters, Wald tests and Wald-based confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W \n It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n"); + fprintf(fichtm, "\n

    The Wald test results are output only if the maximimzation of the Likelihood is performed (mle=1)\n
    Parameters, Wald tests and Wald-based confidence intervals\n
    W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n
    And Wald-based confidence intervals plus and minus 1.96 * W \n
    It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities and its graphs'.\n
    ",optionfilehtmcov); + fprintf(fichtm,"\n"); + fprintf(fichtm, "\n"); + if(nagesqr==1){ + printf(" + age*age "); + fprintf(ficres," + age*age "); + fprintf(ficlog," + age*age "); + fprintf(fichtm, ""); + } + for(j=1;j <=ncovmodel-2;j++){ + if(Typevar[j]==0) { + printf(" + V%d ",Tvar[j]); + fprintf(fichtm, "",Tvar[j]); + }else if(Typevar[j]==1) { + printf(" + V%d*age ",Tvar[j]); + fprintf(fichtm, "",Tvar[j]); + }else if(Typevar[j]==2) { + fprintf(fichtm, "",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]); + } + } + fprintf(fichtm, "\n"); + for(i=1,jk=1; i <=nlstate; i++){ for(k=1; k <=(nlstate+ndeath); k++){ if (k != i) { + fprintf(fichtm, ""); printf("%d%d ",i,k); fprintf(ficlog,"%d%d ",i,k); + fprintf(fichtm, "",i,k); for(j=1; j <=ncovmodel; j++){ - printf("%12.7f W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk], p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk])); - fprintf(ficlog,"%12.7f W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk], p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk])); + wald=p[jk]/sqrt(matcov[jk][jk]); + printf("%12.7f(%12.7f) W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk],sqrt(matcov[jk][jk]), p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk])); + fprintf(ficlog,"%12.7f(%12.7f) W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk],sqrt(matcov[jk][jk]), p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk])); + if(fabs(wald) > 1.96){ + fprintf(fichtm, "", p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk])); jk++; } printf("\n"); fprintf(ficlog,"\n"); + fprintf(fichtm, "\n"); } } } } /* end of hesscov and Wald tests */ + fprintf(fichtm,"
    Model=1+ age+ age*age+ V%d+ V%d*age+ V%d*V%d
    %1d%1d%12.7f
    (%12.7f)
    ",p[jk],sqrt(matcov[jk][jk])); + }else{ + fprintf(fichtm, "
    %12.7f (%12.7f)
    ",p[jk],sqrt(matcov[jk][jk])); + } + fprintf(fichtm,"W=%8.3f
    ",wald); + fprintf(fichtm,"[%12.7f;%12.7f]
    \n"); /* */ fprintf(ficres,"# Scales (for hessian or gradient estimation)\n"); @@ -12640,9 +12962,9 @@ Please run with mle=-1 to get a correct num_filled=sscanf(line,"result:%[^\n]\n",resultline); nresult++; /* Sum of resultlines */ printf("Result %d: result:%s\n",nresult, resultline); - if(nresult > MAXRESULTLINES){ - printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\nYou can use the 'r' parameter file '%s' which uses option mle=0 to get other results. ",MAXRESULTLINES,nresult,rfileres); - fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\nYou can use the 'r' parameter file '%s' which uses option mle=0 to get other results. ",MAXRESULTLINES,nresult,rfileres); + if(nresult > MAXRESULTLINESPONE-1){ + printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\nYou can use the 'r' parameter file '%s' which uses option mle=0 to get other results. ",MAXRESULTLINESPONE-1,nresult,rfileres); + fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\nYou can use the 'r' parameter file '%s' which uses option mle=0 to get other results. ",MAXRESULTLINESPONE-1,nresult,rfileres); goto end; } if(!decoderesult(resultline, nresult)){ /* Fills TKresult[nresult] combination and Tresult[nresult][k4+1] combination values */ @@ -12943,9 +13265,9 @@ Please run with mle=-1 to get a correct for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */ if(i1 != 1 && TKresult[nres]!= k) continue; - printf("\n#****** Result for:"); - fprintf(ficrest,"\n#****** Result for:"); - fprintf(ficlog,"\n#****** Result for:"); + printf("\n# model %s \n#****** Result for:", model); + fprintf(ficrest,"\n# model %s \n#****** Result for:", model); + fprintf(ficlog,"\n# model %s \n#****** Result for:", model); for(j=1;j<=cptcoveff;j++){ printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);