steckt der fehler bei covbta in glsfit? (muesst auch in lsfit sein,

stuertzt auch mit method="ols" ab). covbta hat zumindest keinen
sinnvollen inhalt nach! glsfit

devel/fresh/.Rhistory

+gc()
+ls()
+leman.bk.0 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.0,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=100, ny=100, trend=0, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
+library(baykrig)
+leman.bk.0 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.0,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=100, ny=100, trend=0, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
+plot(leman.bk.0)
+leman.prior.0
+leman.bk.0 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.0,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=50, ny=50, trend=0, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
+plot(leman.bk.0)
+source("leman-example.R")
+library(baykrig)
+leman.bk.0 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.0,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=50, ny=50, trend=0, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
+leman.bk.0 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.0,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=100, ny=100, trend=0, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
+plot(leman.bk.0)
+dev.print("leman-0.ps")
+dev.print(file="leman-0.ps")
+library(baykrig)
 library(baykrig)
 library(baykrig)
 library(baykrig)
 library(baykrig)
-leman.bk1 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.1,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=10, ny=10, trend=1, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
 library(baykrig)
-leman.bk1 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.1,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=10, ny=10, trend=1, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
-plot(leman.bk1)
-leman.bk1$z
-leman.bk1$snb
-leman.bk1$
-q()
-source("leman-example.R")
-leman.prior.0<-empirical.prior(leman.83,cadpbm~1,leman.88.vmsph,duplicate="mean",name.x="x",name.y="y")
-leman.prior.0<-empirical.prior(leman.83,cadpbm~1,leman.88.vmsph,duplicate="mean",namx="x",namy="y")
-leman.prior.0<-empirical.prior(leman.78,cadpbm~1,leman.88.vmsph,prior=leman.prior,duplicate="mean",namx="x",namy="y"))
-leman.prior.0<-empirical.prior(leman.78,cadpbm~1,leman.88.vmsph,prior=leman.prior,duplicate="mean",namx="x",namy="y")
-save.image()
-leman.bk.0 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.0,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=50, ny=50, trend=0, rsearch = 10, extrap = T,border=leman.bank, duplicate="mean")
-leman.prior.0
-rm(leman.prior.0)
-leman.prior.0<-empirical.prior(leman.83,cadpbm~1,leman.88.vmsph,duplicate="mean",namx="x",namy="y")
-leman.prior.0
-leman.prior.0<-empirical.prior(leman.78,cadpbm~1,leman.88.vmsph,prior=leman.prior.0,duplicate="mean",namx="x",namy="y")
-save.image()
-leman.prior.0
-leman.bk.0 <- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior.0,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=50, ny=50, trend=0, rsearch = 10, extrap = T,border=leman.bank, duplicate="mean")
-str(leman.bk.0)
-save.image()
-leman.bk.0$z
-gc()
 q()
+library(baykrig)
+leman.bk<- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=100, ny=100, trend=1, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
+library(baykrig)
+leman.bk<- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=100, ny=100, trend=1, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")
+library(baykrig)
+library(baykrig)
+leman.bk<- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=100, ny=100, trend=1, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean",method="ols")
+library(baykrig)
+leman.bk<- bk.grid(point = leman.88.pt , at = "cadpbm", prior=leman.prior,var.mod.obj = leman.88.vmsph, xsw=min(leman.bank$x),ysw=min(leman.bank$y), xne=max(leman.bank$x), yne=max(leman.bank$y), nx=100, ny=100, trend=1, rsearch = 10, extrap = F,border=leman.bank, duplicate="mean")

src/bk.f

@@ -246,17 +246,21 @@ c      call imatpr(do0,n0,1,n0,name,dbglvl)
       IF (NDO.EQ.0) THEN 
          RETURN
       END IF
+
       
 c     determine search neighbourhood
       CALL SRCHNB(LON0,LAT0,DO0,INDDO,N0,NDO,LON,LAT,N,NS,
      .            RSEARCH,NSEARCH,NSMIN,NSMAX,
      .            INDSRT,INDSNB,INDSNW,DIST)
+
 c     prepare the design matrix
       CALL DESIGN(LON,LAT,N,INDSNB,NS,LON0,LAT0,N0,INDDO,NDO,
      .            FWORK,LDFWRK,F0WORK,LDF0WK,NTREND,TREND,IERR)
+
 c     make a copy in fwrk2
       CALL DESIGN(LON,LAT,N,INDSNB,NS,LON0,LAT0,N0,INDDO,NDO,
      .            FWRK2,LDFWRK,F0WORK,LDF0WK,NTREND,TREND,IERR)
+
 c      CALL DSUBMM(FWORK,NS,NTREND,1,1,NS,NTREND,LDFWRK,FWRK2,LDFWRK,1)
 c     extract appropriate parts of covariance matrix and Z
       DO 10 I=1,NS
@@ -267,6 +271,7 @@ c     extract appropriate parts of covariance matrix and Z
  10   CONTINUE
 
 
+
 c     do (generalised) least squares fit in search neighbourhood
 c      CALL GLSFIT(FWORK,N,NTREND,LDFWRK,ZSRNB,CVSRNB,LDCOV,BETA,U,
 c     .            COVBTA,CHLUP,CMINV,WORK1,WORK,LWORK,IPVT,IWORK,IERR)
@@ -286,7 +291,7 @@ c     .            COVBTA,CHLUP,CMINV,WORK1,WORK,LWORK,IPVT,IWORK,IERR)
          END IF
       ELSE
 c         CALL GLSFIT(FWORK,FWRK2,NS,NTREND,LDFWRK,ZSRNB,CVSRNB,LDCOV,
-         CALL GLSFIT(FWORK,FWork,NS,NTREND,LDFWRK,ZSRNB,CVSRNB,LDCOV,
+         CALL GLSFIT(FWORK,FWRK2,NS,NTREND,LDFWRK,ZSRNB,CVSRNB,LDCOV,
      .               BETA,ERRBTA,DEV,ERRDEV,
      .               COVBTA,LDCVBT,SGSQR,
      .               CHLUP,LDCLUP,CMINV,LDCINV,
@@ -304,7 +309,10 @@ c      call matpr(name,beta,ntrend,1,ntrend,1)
 c      name="covbta\0"
 c      call matpr(name,covbta,ntrend,ntrend,ntrend,1)
       END IF
-cccc      goto 123
+
+      goto 123
+
+
 
 c     merge priors with search neighbourhood:
 c     average all prior guesses with estimation in search neighbourhood
@@ -519,7 +527,7 @@ c      name="z0"
 c      call matpr(z0,n0,1,n0,name,dbglvl)
       write(*,*)"z: ",z0(1)
       IWORK(1)=NS
-cccc 123  continue
+ 123  continue
       RETURN
       END