Re: Analyzing bootstrap output in R

From: Saik Urien Svp Date: November 26, 2008 technical Source: mail-archive.com
Kellie As far as I can see from your input, the problem arises because some rows must have NA or blank or no values. I suspect this can be due to the theta.se. A simple way to fix the problem could be to delete the $COVARIANCE statement. Saik
Quoted reply history
----- Original Message ----- From: "Kellie Turner" <[EMAIL PROTECTED]> To: <[email protected]> Sent: Wednesday, November 26, 2008 9:45 AM Subject: [NMusers] Analyzing bootstrap output in R > From: Kellie Turner > Sent: 25 November 2008 15:38 > To: [email protected] > Subject: Analyzing bootstrap output in R > > Dear NONMEM users, > I am trying to analyze in R the results of a bootstrap run of a model. The bootsrap runs fine with NONMEM VI level 1.0, implemented in PDxPOP 3.0. > However, when I try to view the output in R, I get this error message: > Error in data.frame (est=c(unlist(theta)), se= c(unlinst(theta.se))): arguments imply differing number of rows:7, 0. > > The data is from 2 courses,and the problem seems to be with including both courses, as I have no problems when I include only 1 course. I have tried going back to the base model (no covariates, no IOV, no correlation between CL and V), and I get the same error message when data from both courses are included. > Does anyone know how I can fix this problem so that I can view the bootstrap results in R? > Thanks in advance for your assistance, > Kellie > > > Here is the control file > $INPUT C ID TIME DV AMT RATE EVID GFR GT1 GT2 WT OCC > $DATA bootstrap.csv IGNORE=C > $SUBROUTINES ADVAN3 TRANS4 > $PK > HET=0 > HOM=0 > UNK=0 > IF(GT1.EQ.2) HET=1 > IF(GT1.EQ.3) HOM=1 > IF(GT1.EQ.0) UNK=1 > > OCC1=0 > OCC2=0 > IF(OCC.EQ.0) OCC1=1 > IF(OCC.EQ.1) OCC2=1 > > IF (AMT.GT.0) THEN > TDOS=TIME > TAD= 0.0 > ENDIF > IF (AMT.EQ.0) TAD=TIME-TDOS > > TVCL=THETA(1)*(1-(THETA(5)*HET + 2*THETA(5)*HOM))*(1-THETA(6)*UNK) > CL=TVCL*EXP(ETA(1)+OCC1*ETA(5)+OCC2*ETA(6)) > TVV1=THETA(2) + THETA(7)*(WT/70) > TVQ=THETA(3) > TVV2=THETA(4) > V1=TVV1*EXP(ETA(2)+OCC1*ETA(7)+OCC2*ETA(8)) > Q=TVQ*EXP(ETA(3)) > V2=TVV2*EXP(ETA(4)) > S1=V1 > AUC=AMT/CL > SID=ID > > $ERROR > DEL=0 > IF (F.EQ.0) DEL=1 > W=F+DEL > Y=F+W*EPS(1) > IPRED=F > IRES=DV-IPRED > IWRES=IRES/W > $THETA (0,.05,2) ; [CL] > $THETA (5,10,20) ; [V1] > $THETA (0,.2,2) ; [Q] > $THETA (3,9,30) ; [V2] > $THETA (0,.05,.5) ; [GT1] > $THETA (0,.05,.5) ; [UNK GT1] > $THETA (0,5,20) ; [WT] > $OMEGA BLOCK (2) > 0.5 ;[P] (1,1) > 0.1 ;[F] (1,2) > 0.5 ;[P] (2,2) > $OMEGA 0 FIXED ;[P] INTER IND VAR IN Q > $OMEGA 0.1 ;[P] INTERIND VAR IN V2 > $OMEGA BLOCK (1).2 ;[F] > $OMEGA BLOCK (1) SAME ;[F] > $OMEGA BLOCK (1).2 ;[F] > $OMEGA BLOCK (1) SAME ;[F] > $SIGMA 0.05 ;[P] PROPORTIONAL COMPONENT > > ;$MSFI > $EST METHOD=0 SIGDIGITS=4 POSTHOC NOABORT MSF=boot-275.msf > $COVARIANCE > $TABLE FILE=boot-275.TAB ID TIME CL DV V1 V2 Q GT1 WT IPRED PRED ETA1 ETA2 ETA3 ETA4 ETA5 ETA6 ETA7 ETA8 NOPRINT ONEHEADER > $TABLE ID NOPRINT ONEHEADER FILE=patabboot-275 > $TABLE ID NOPRINT ONEHEADER FILE=cotabboot-275 > $TABLE ID NOPRINT ONEHEADER FILE=catabboot-275 > $TABLE ID TIME IPRED IWRES NOPRINT ONEHEADER FILE=sdtabboot-275 >
Nov 26, 2008 Kellie Turner Analyzing bootstrap output in R
Nov 26, 2008 Saik Urien Svp Re: Analyzing bootstrap output in R