Appendix

Standard Pharmacokinetic Models and Parameters

This appendix lists the ADVAN routines for the most commonly-used pharmacokinetic models, along with their compartments and basic and additional PK parameters.

ADVAN1 (One Compartment Linear Model)

Compartments:

• 1 = Central
• 2 = Output

Basic and additional PK parameters:

Parameter	Description
K	Rate constant of elimination
S1	Scale for central compartment
S2	Scale for output compartment
F1	Bioavailability for central compartment
F0	Output Fraction

ADVAN2 (One Compartment Linear Model with First Order Absorption)

Compartments:

• 1 = Depot
• 2 = Central
• 3 = Output

Basic and additional PK parameters:

Parameter	Description
KA	Absorption rate constant
K	Rate constant of elimination
S1	Scale for depot compartment
S2	Scale for central compartment
S3	Scale for output compartment
F1	Bioavailability for depot compartment
F2	Bioavailability for central compartment
F0	Output Fraction

ADVAN3 (Two Compartment Linear Model)

Compartments:

• 1 = Central
• 2 = Peripheral
• 3 = Output

Basic and additional PK parameters:

Parameter	Description
K	Rate constant of elimination
K12	Rate constant from central to peripheral
K21	Rate constant from peripheral to central
S1	Scale for central compartment
S2	Scale for peripheral compartment
S3	Scale for output compartment
F1	Bioavailability for central compartment
F2	Bioavailability for peripheral compartment
F0	Output Fraction

ADVAN4 (Two Compartment Linear Model with First Order Absorption)

Compartments:

• 1 = Depot
• 2 = Central
• 3 = Peripheral
• 4 = Output

Basic and additional PK parameters:

Parameter	Description
KA	Absorption rate constant
K	Rate constant of elimination
K23	Rate constant from central to peripheral
K32	Rate constant from peripheral to central
S1	Scale for depot compartment
S2	Scale for central compartment
S3	Scale for peripheral compartment
S4	Scale for output compartment
F1	Bioavailability for depot compartment
F2	Bioavailability for central compartment
F3	Bioavailability for peripheral compartment
F0	Output Fraction

ADVAN10 (One Compartment Model with Michaelis-Menten Elimination)

Compartments:

• 1 = Central
• 2 = Output

Basic and additional PK parameters:

Parameter	Description
VM	Maximum Rate
KM	Michaelis Constant
S1	Scale for central compartment
S2	Scale for output compartment
F1	Bioavailability for central compartment
F0	Output Fraction

ADVAN11 (Three Compartment Linear Model)

Compartments:

• 1 = Central
• 2 = Peripheral 1
• 3 = Peripheral 2
• 4 = Output

Basic and additional PK parameters:

Parameter	Description
K	Rate constant of elimination
K12	Rate constant from central to peripheral 1
K21	Rate constant from peripheral 1 to central
K13	Rate constant from central to peripheral 2
K31	Rate constant from peripheral 2 to central
S1	Scale for central compartment
S2	Scale for peripheral 1 compartment
S3	Scale for peripheral 2 compartment
S4	Scale for output compartment
F1	Bioavailability for central compartment
F2	Bioavailability for peripheral 1 compartment
F3	Bioavailability for peripheral 2 compartment
F0	Output Fraction

ADVAN12 (Three Compartment Linear Model with First Order Absorption)

Compartments:

• 1 = Depot
• 2 = Central
• 3 = Peripheral 1
• 4 = Peripheral 2
• 5 = Output

Basic and additional PK parameters:

Parameter	Description
KA	Absorption rate constant
K	Rate constant of elimination
K23	Rate constant from central to peripheral 1
K32	Rate constant from peripheral 1 to central
K24	Rate constant from central to peripheral 2
K42	Rate constant from peripheral 2 to central
S1	Scale for depot compartment
S2	Scale for central compartment
S3	Scale for peripheral 1 compartment
S4	Scale for peripheral 2 compartment
S5	Scale for output compartment
F1	Bioavailability for depot compartment
F2	Bioavailability for central compartment
F3	Bioavailability for peripheral 1 compartment
F4	Bioavailability for peripheral 2 compartment
F0	Output Fraction

Alternative Parameterizations

This appendix describes the alternative parameterizations (TRANS subroutines) available in the PREDPP Library. Each TRANS routine translates a specified set of alternative parameters into the microconstants expected by PREDPP.

ADVAN1 TRANS2

Alternative Parameters:

Parameter	Description
CL	Clearance
V	Volume of distribution

Reparameterization line:

K=CL/V

ADVAN2 TRANS2

Alternative Parameters:

Parameter	Description
CL	Clearance
V	Volume of distribution
KA	Absorption rate

Reparameterization lines:

K=CL/V
KA=KA

ADVAN3 TRANS3

Alternative Parameters:

Parameter	Description
CL	Clearance
V	Central Volume
Q	Intercompartmental clearance
VSS	Volume of distribution at steady state

Reparameterization lines:

K=CL/V
K12=Q/V
K21=Q/(VSS-V)

ADVAN3 TRANS4

Alternative Parameters:

Parameter	Description
CL	Clearance
V1	Central volume
Q	Intercompartmental clearance
V2	Peripheral volume

Reparameterization lines:

K=CL/V1
K12=Q/V1
K21=Q/V2

ADVAN3 TRANS5

Alternative Parameters:

Parameter	Description
AOB	A/B
ALPHA	alpha
BETA	beta

Reparameterization lines:

K21=(AOB*BETA+ALPHA)/(AOB+1)
K=ALPHA*BETA/K21
K12=ALPHA+BETA-K21-K

ADVAN3 TRANS6

Alternative Parameters:

Parameter	Description
ALPHA	alpha
BETA	beta
K21	Rate constant from peripheral to central

Reparameterization lines:

K=ALPHA*BETA/K21
K12=ALPHA+BETA-K21-K
K21=K21

ADVAN4 TRANS3

Alternative Parameters:

Parameter	Description
CL	Clearance
V	Central Volume
Q	Intercompartmental clearance
VSS	Volume of distribution at steady state
KA	Absorption rate

Reparameterization lines:

K=CL/V
K23=Q/V
K32=Q/(VSS-V)
KA=KA

ADVAN4 TRANS4

Alternative Parameters:

Parameter	Description
CL	Clearance
V2	Central volume
Q	Intercompartmental clearance
V3	Peripheral volume
KA	Absorption rate

Reparameterization lines:

K=CL/V2
K23=Q/V2
K32=Q/V3
KA=KA

ADVAN4 TRANS5

Alternative Parameters:

Parameter	Description
AOB	A/B
ALPHA	alpha
BETA	beta
KA	Absorption rate

Reparameterization lines:

K32=(AOB*BETA+ALPHA)/(AOB+1)
K=ALPHA*BETA/K32
K23=ALPHA+BETA-K32-K
KA=KA

ADVAN4 TRANS6

Alternative Parameters:

Parameter	Description
ALPHA	alpha
BETA	beta
K32	Rate constant from peripheral to central
KA	Absorption rate

Reparameterization lines:

K=ALPHA*BETA/K32
K23=ALPHA+BETA-K32-K
K32=K32
KA=KA

ADVAN11 TRANS4

Alternative Parameters:

Parameter	Description
CL	Clearance
V1	Central volume
Q2	Intercompartmental clearance 1
V2	Peripheral volume 1
Q3	Intercompartmental clearance 2
V3	Peripheral volume 2

Reparameterization lines:

K=CL/V1
K12=Q2/V1
K21=Q2/V2
K13=Q3/V1
K31=Q3/V3

ADVAN11 TRANS6

Alternative Parameters:

Parameter	Description
ALPHA	alpha
BETA	beta
GAMMA	gamma
K21	Rate constant from peripheral 1 to central
K31	Rate constant from peripheral 2 to central

Reparameterization lines (approximate form; see Users Guide VI for full expressions):

K=ALPHA*BETA*GAMMA/(K21*K31)
K13=(V2+K31*K31-K31*V1-K*K21)/(K21-K31)
K12=V1-K-K13-K21-K31

where V1, V2 are computed from ALPHA, BETA, GAMMA.

ADVAN12 TRANS4

Alternative Parameters:

Parameter	Description
CL	Clearance
V2	Central volume
Q3	Intercompartmental clearance 1
V3	Peripheral volume 1
Q4	Intercompartmental clearance 2
V4	Peripheral volume 2
KA	Absorption rate

Reparameterization lines:

K=CL/V2
K23=Q3/V2
K32=Q3/V3
K24=Q4/V2
K42=Q4/V4
KA=KA

ADVAN12 TRANS6

Alternative Parameters:

Parameter	Description
ALPHA	alpha
BETA	beta
GAMMA	gamma
K32	Rate constant from peripheral 1 to central
K42	Rate constant from peripheral 2 to central
KA	Absorption rate

Reparameterization lines (approximate form; see Users Guide VI for full expressions):

K=ALPHA*BETA*GAMMA/(K32*K42)
K24=(V3+K42*K42-K42*V2-K*K32)/(K32-K42)
K23=V2-K-K24-K32-K42
KA=KA

where V2, V3 are computed from ALPHA, BETA, GAMMA.

NM-TRAN Control Records

The following is an alphabetic list of NM-TRAN control records. See Guide VIII, On-line Help, or On-line HTML for the options and for more information. See Appendix 4 for the corresponding NONMEM control records (FCON).

• $ABBREVIATED
• $AES
• $AESINIT
• $ANNEAL
• $BIND
• $CHAIN
• $CONTR
• $COVARIANCE
• $DATA
• $DEFAULTS
• $DES
• $DESIGN
• $ERROR
• $ESTIMATION
• $ETAS
• $FORMAT
• $INDEX
• $INFN
• $INPUT
• $LEVEL
• $LDFOLD
• $MIX
• $MODEL
• $MSFI
• $NONPARAMETRIC
• $OMEGA
• $OMEGAP
• $OMEGAPD
• $OLKJDF
• $OMIT
• $OVARF
• $PHIS
• $PK
• $PRED
• $PRIOR
• $PROBLEM
• $RCOV
• $RCOVI
• $SCATTER
• $SIGMA
• $SIGMAP
• $SIGMAPD
• $SIMULATION
• $SIZES
• $SLKJDF
• $SUBROUTINES
• $SUPER
• $SVARF
• $TABLE
• $THETA
• $THETAI
• $THETAP
• $THETAPV
• $THETAR
• $TTDF
• $TOL
• $WARNING

NONMEM Control Records (FCON)

The following is a list of NONMEM control records and options. These are generated by NM-TRAN in a file called FCON. They are listed in the order that they appear in FCON.

Records marked with * may be continued. The record name, e.g., INDX, is not repeated on continuation(s). Any time a field may contain 0, it may contain blanks instead, which are read as 0. Constants SD, PD, LVR are from SIZES. Constants INF, INTBIG, INTSMALL indicate the largest floating point number, the largest integer value, and smallest integer value, respectively, that can be represented in the computer’s architecture. (In some cases, e.g., the BOOTSTRAP option of SIMULATION, it means an integer of up to 11 digits).

Note

This appendix presents an overview of FCON record structure. For complete, byte-level details of each field in each record, consult NONMEM Users Guide Part VIII (the Help Guide) or the on-line HTML documentation.

FILE record

(FILE) (A4,4X,A72)

Specifies the name of a file stream (up to 72 characters), or NULL for no file stream.

SUPER record

(SUPR) (A4,4X,I4,I8,I4)

Fields specify:

1. Number of problems in the superproblem (1-9999)
2. Number of iterations of the superproblem (2-9999)
3. Whether input information is printed for first problem only (0) or all problems (1)

PROBLEM record

(PROB) (A4,4X,A72)

Contains the problem heading (up to 72 characters).

FORMAT record

(FMTN) (A4,I4)

Specifies significant digits for output file in 1PE (positive) or 1PG (negative) format. Valid values: -23 to -3, 3 to 23.

LDFOLD record

(LDFO) — no options. Present if the $LDFOLD record was used.

QLOGMAX record

(QLMX) (A4,I4)

Maximum exponent for PEXP() (in PROTECT.f90), and for IMP sample scaling. Default is 300; valid range 300-350.

DATA record

(DATA) (A4,4X,8I4,I12,I4,I2)

Describes characteristics of the input data file:

1. Whether data set is embedded (0), in a separate file (1), or re-used from previous problem (-1)
2. Whether FORTRAN unit is rewound (0=no, 1=yes)
3. Number of data records (low-order digits)
4. Number of data items per data record (1 to PD)
5. Data checkout mode (0=no, 1=yes)
6. Number of data records (high-order digits). Total = Field 6 × 10000 + Field 3.
7. Simulation REWIND from $SIMULATION record (0=NOREWIND, 1=REWIND)
8. NOSUPRESET (0) or SUPRESET (1, default)
9. REPL value (0 to INTBIG)
10. PRED_IGNORE_DATA flag (0 or 1)
11. NOFDATACSV flag (0 or 1)

ITEM record

(ITEM) (A4,4X,18I4)

Describes data item positions:

1. Index of ID data item
2. Index of DV data item
3. Index of MDV data item
4. Number of data item indices in INDXS
5. User-supplied labels flag (0=no, 1=yes)
6. Standard vs nonstandard labels for PRED, RES, WRES (0=standard, 1=nonstandard)
7. Index of L2 data item 8-10. Indices of data items specified in CONTR record
8. Number of user-supplied labels for tables/scatters
9. Index of MRG_ data item
10. Index of RAW_ data item
11. Number of items on OMIT record
12. Index of RPT_ data item

INDEX record

(INDX)* (A4,4X,18I4)

Lists elements of INDXS (1 to PD each).

XVID record

(XVID) (A4,4X,18I4)

Fields 1-5 are indices for XVID1 through XVID5 (0 to PD each).

MSDT record

(MSDT) (A4,4X,20(1PE22.14E3,1X))

Fields 1-20 contain MISDAT(1) through MISDAT(20) values.

LABEL record

(LABL)* (A4,X2,A74)

Contains a comma-delimited list of labels starting at position 6, with an unlimited number of continuation records. Each label is right-adjusted in a field of SD characters. The order is:

1. Label of 1st data item
2. Label of 2nd data item, etc.
3. Label for PRED, RES, WRES (if ITEM(6)=1)
4. Labels for user-displayed variables

Additional LABEL record

(LBW1)* (A4,X2,A74)

Contains labels for the additional diagnostic items, in this default order:

• IWRS, IPRD, IRS
• NPRED, NRES, NWRES
• NIWRES, NIPRED, NIRES
• CPRED, CRES, CWRES
• CIWRES, CIPRED, CIRES
• PREDI, RESI, WRESI
• IWRESI, IPREDI, IRESI
• CPREDI, CRESI, CWRESI
• CIWRESI, CIPREDI, CIRESI
• EPRED, ERES, EWRES
• EIWRES, EIPRED, EIRES
• NPDE, ECWRES, NPD
• OBJI

LABEL records for THETA, ETA, EPS

• LTHT)* (A4,X4,A72) — Symbolic names for elements of THETA
• LETA)* (A4,X4,A72) — Symbolic names for elements of ETA
• LEPS)* (A4,X4,A72) — Symbolic names for elements of EPS
• LRES) (A4,X1,A5) — RESIDUAL labels

These support label substitution in the NONMEM report file.

For example, if the NM-TRAN control file contains:

$ABBR REPLACE THETA(KA,K,CL)=THETA(1 TO 3)
$ABBR REPLACE ETA(CL)=ETA(3),ETA(V)=ETA(5)

Then the generated LABEL records are:

LTHT 1=THETA(KA),2=THETA(K),3=THETA(CL)
LETA 3=ETA(CL),5=ETA(V)
LEPS

OMIT record

(OMIT)* (A4,4X,18I4)

Lists numbers of data items omitted from template matching.

FORMAT record

(FORM) (A4,4X,A72/A80)

Contains the FORTRAN format specification (begins on first continuation record).

FIND record

(FIND) (A4,4X,18I4)

Controls Model Specification File (MSFI) handling:

• Fields 3, 4, 5, 6: MSFI options including whether to read MSF, rescale, ONLYREAD, MSFTEST
• Field 7: MSFI new flag

MSF Version record

(MSFV) (A4,X4,A72)

Contains the MSF version (e.g., MSFV 7.2.0).

INITIAL VALUES records for ETA and PHI

(ETA)* (A4,I4,...) and (PHI)* (A4,I4,...)

Either specify a file name (with FILE), or list initial values directly for etas/phis.

Initial STRUCTURE record

(STRC) (A4,4X,18I4)

Fields describe problem structure:

1. Length of THETA (0-999)
2. Dimension of OMEGA (0-999)
3. Dimension of SIGMA (0-999)
4. Reserved
5. OMEGA block set partition flag
6. OMEGA diagonal constraint flag
7. Number of block sets for OMEGA 8-9. Same for SIGMA 10-11. Reserved
8. No default THETA boundary test flag
9. No default OMEGA boundary test flag
10. No default SIGMA boundary test flag

STRUCTURE records for OMEGA and SIGMA

(STRC)* (A4,4X,18I4)

Describe block structure: size and dimension of each block set.

Records for TTDF, OVARF, SVARF, OLKJDF, SLKJDF

(A4,4X,I12) for counts, followed by comma-delimited value lists.

THETA CONSTRAINT record

(THCN) (A4,4X,18I4)

Fields specify:

1. THETA unconstrained (0) or constrained (1)
2. Size of initial estimate search (0 for default, 1-9999 otherwise)
3. PRED error handling: ABORT (0), NOABORT (1), NOABORTFIRST (2)

THETA record

(THTA)* (A4,4X,(comma-delimited list))

Initial estimates of \(\theta_1\), \(\theta_2\), etc. Blank indicates NONMEM is to obtain the initial estimate.

THETA uninteresting for DESIGN

(THTU)* (A4,4X,(comma-delimited list))

Marks each theta as uninteresting (1) or interesting (0) for design.

LOWER and UPPER BOUND records

• (LOWR)* (A4,4X,(comma-delimited list)) — Lower bounds for each \(\theta\)
• (UPPR)* (A4,4X,(comma-delimited list)) — Upper bounds for each \(\theta\)

DIAGONAL record for OMEGA or SIGMA

(DIAG)* (A4,1X,A1,1X,A1,(comma-delimited list))

First character positions specify:

• Variance (0) vs standard deviation (1) of diagonals
• Not fixed (blank), Fixed (1), NONMEM obtains initial estimate (2), Uninteresting for DESIGN (9)

Followed by initial estimates of diagonal elements.

BLOCK SET record for OMEGA or SIGMA

(BLST)* (A4,1X,A1,1X,A1,(comma-delimited list))

First character position specifies parameterization:

• 0 (blank): Diagonal variance, off-diagonal covariance (default)
• 1: Diagonal SD, off-diagonal covariance (STANDARD)
• 2: Diagonal variance, off-diagonal correlation (CORRELATION)
• 3: Diagonal SD, off-diagonal correlation (STANDARD CORRELATION)
• 4: Cholesky format (CHOLESKY)

Second character position specifies fixed/not-fixed status. Initial estimates follow in symmetric enumeration order.

DESIGN record

(OPTD) (A4,4X,I2,I3,2I2,7I6,I12,I2,E23.16,I6,I3,2I6,I3)

Controls design/evaluation options including OPTDTYPE (FEDOROV, NELDER, STGR, RS, DISCRETE, etc.), OFVTYPE, SEED, CLOCKSEED, GROUPSIZE, MODE, DATASIM, DOPTCOVCOLS, FIMDIAG, VARCROSS.

SIMULATION record

(SIML) (A4,4X,I2,I3,I2,I2,I13,5I2,I13,4(I6),1X,A16)

Main simulation control fields:

1. Simulation Step implemented (0) or not (1)
2. Number of random sources
3. Eta/eps changes per record (0) or per individual (1)
4. Number of subproblems
5. Only simulation step (1) or full computations (0)
6. PRED derivatives: none (0), first (1 REQUESTFIRST), second (2 REQUESTSECOND)
7. PREDICTION (0) vs NOPREDICTION (1)
8. TRUE source: INITIAL (0), FINAL (1, with MSFI), PRIOR (2)
9. REPLACE (0) or NOREPLACE (1)
10. BOOTSTRAP: disabled (0), full sample (-1), specific count (1 to INTBIG) 11-12. STRAT, STRATF data column numbers
11. SOURCE_EPS
12. TTDF
13. RANMETHOD

Additional records provide PARAFILE information.

SOURCE record

(SORC) (A4,4X,2A12,I4)

Fields:

1. First seed (-1 to 21474836447)
2. Second seed (0 to 21474836447)
3. Distribution type: NORMAL (0), UNIFORM (1), NONPARAMETRIC (2)

DEFAULT record

(DFLT) (A4,4X,I4)

NOSUB option of $DEFAULTS record (-1 to 1).

INITIAL VALUES records for RCOV and RCOVI

(RCOV)* and (RCVI)* specify filenames and additional options (DELIM, TBLN).

CHAIN record

(CHN)

Complex multi-line record controlling chain estimation methods with fields for CTYPE, SEED, ISAMPLE, NSAMPLE, IACCEPT, DF, ISAMPEND, SELECT, NOTITLE, NOLABEL, DFS, RANMETHOD.

Additional records provide FILE, FORMAT, ORDER, parallelization options.

LEVEL record

(OLEV)* (A4,X4,A20,A52)

Specifies random effect level data item type and description.

ANNEAL record

(ANNL) (A4,X4,A6,A6)

Eta number (1 to LVR) and starting omega value for annealing.

ESTIMATION record

(ESTM) (A4,4X,18I4)

Main estimation fields:

1. Estimation Step implemented (0) or not (1)
2. Maximum function evaluations (low-order digits)
3. Number of significant figures required (1-8)
4. Iteration summarization interval (n or 0)
5. REPEAT option (0 or 1)
6. MSF output flag (0 or 1)
7. FO (0) or Conditional (1) method
8. POSTHOC option (0 or 1)
9. INTERACTION option (0 or 1)
10. Laplacian method (0 or 1)
11. PRED error recovery: ABORT (0), NOABORT (1), NOHABORT (2)
12. Computation speed: NOSLOW (0), SLOW (1), SLOW=2 (2), FAST (3)
13. NOCENTER (0) or CENTER (1)
14. FO model usage flag
15. NUMERICAL option
16. LIKELIHOOD flag: prediction (0), LIKELIHOOD (1), -2LOGLIKELIHOOD (2)
17. Hybrid method: not used (0), number of etas fixed (1-99)
18. Stieltjes method flag

ESTIMATION continuation records

The ESTIMATION record extends with many continuation lines specifying options such as REPEAT2, ETABARCHECK, SORT, SIGL/SIGLO defaults, and many others.

BAYES ESTIMATION record

(BEST) — Very extensive multi-line record (lines 1-16) controlling Bayesian and EM estimation methods, including:

• Method selection (BAYES METHOD: -1-16, covering FO/FOCE/Laplace, DIRECT, BAYES, ITS, SAEM, IMP, IMPMAP, CHAIN)
• Sample sizes (PSAMPLE_M1/M2/M3, OSAMPLE_M1/M2/M3, ISAMPLE, NSAMPLE, NBURN)
• Acceptance ratios (PACCEPT, OACCEPT, IACCEPT)
• Control parameters (EONLY, SEED, NOPRIOR, CTYPE, CITER, CALPHA, CINTERVAL)
• MAP options (MAPITER, MAPINTER, OPTMAP, ETADER, MCETA, NONINFETA)
• NUTS options (BAYES_METHOD, MADAPT, KAPPA, NUTS_GAMMA, NUTS_DELTA, NUTS_TYPE, NUTS_MASS, NUTS_TRANSFORM, NUTS_EPARAM, NUTS_OPARAM, NUTS_SPARAM, NUTS_STEPITER, NUTS_STEPINTER, NUTS_TEST, NUTS_INIT, NUTS_BASE, NUTS_TERM, NUTS_MAXDEPTH, NUTS_CHOLBND, NUTS_REG)
• Additional (GRDQ, MASSRESET, KNUTHSUMOFF, SADDLE_RESET, SADDLE_HESS, THIN, ETASAMPLES, PHITYPE, LEVSWITCH, TBLN, BAYES_PHI_STORE, BIONLY, BOOTDATA, NOLEVCENTER, LEVOBJTYPE, EVALSHRINK)

Additional records for ESTIMATION

• BFIL — FILE (up to 256 characters)
• BDLM — FORMAT/DELIM
• BMUM — MUM
• BGRD — GRD
• ORDR — ORDER
• PFIL — PARAFPRINT, PARAFILE
• FFIL — FPARAFPRINT, FPARAFILE

ZERO record

(ZERO)* (A4,4X,18I4)

For the HYBRID method, specifies which etas are fixed to 0 vs receive conditional estimates.

GRID record

(GRID) (A4,4X,9A8)

Specifies GRID parameters (nr, ns, r0, r1) for the Stieltjes method.

NONPARAMETRIC record

(NONP) (A4,4X,5I4,I12,I12,I6,I6,I2,I12)

Controls nonparametric analysis:

1. Conditional (0) or unconditional (1) implementation
2. Use MSF input (0) vs recompute (1)
3. Marginal cumulatives (0) vs CNPE ETAS (1)
4. MSF output flag
5. BOOTSTRAP flag
6. EXPAND/NPSUPPE mode
7. NPSUPP(E) value 8-9. STRAT, STRATF data column numbers
8. NPESTIM flag
9. NPMAXITER value

Additional NFIL record: PARAFPRINT, PARAFILE.

COVARIANCE record

(COVR) (A4,4X,18I4)

Main covariance step fields:

1. Covariance Step: conditional (0), unconditional (1), not implemented (2)
2. Matrix: default R⁻¹SR⁻¹ (0), R⁻¹ (1), S⁻¹ (2)
3. Print R matrix flag
4. Print S matrix flag
5. Print eigenvalues flag
6. Special computation with recursive PRED
7. Compressed format flag
8. Normal method or slower SLOW method
9. FAST analytical derivative method (FOCE only, value 3)

Additional COVARIANCE records

(COVT) provides detailed options including SIGL, TOL, SIGLO, ATOL, NOFCOV, COVRESUME, SIR (sampling importance resampling) parameters, PRECOND, PFCOND, PRETYPE, FPOSDEF, THBND, SIRTHBND, POSDEF, PRECONDS.

A second COVT line covers SIRPRINT, CHOLROFF, KNUTHSUMOFF, SIRDF, IACCEPT values, SIR_CAPCORR, SIRMAXWT, SIRMINWT, RANMETHOD.

Additional records: CPAR (FORMAT, FILE), SPAR (PARAFPRINT, PARAFILE, SIRPARAFPRINT, SIRPARAFILE).

Initial TABLE record

(TABL) (A4,4X,18I4)

Fields:

1. Table Step: conditional (0), unconditional (1), not implemented (2)
2. Number of tables (1-10)

Additional PPAR record provides PARAFPRINT, PARAFILE.

Individual TABLE record

(TABL) (A4,1X,3A1,I4,9(I6,I2))

Specifies items to include in each table, with sort codes and exclude-by flags.

Individual TABLE option record

Options include:

• Record selection: FIRSTONLY (1), LASTONLY (2), FIRSTLASTONLY (3)
• NOPRINT (1) vs printed (2)
• Header format: default (0), ONEHEADER (1), NOTITLE (4), NOLABEL (8), combinations (5, 9, 14)
• FORWARD option (0=start, 1=forward)
• NOAPPEND option
• SEED, ESAMPLE, RANMETHOD, WRESCHOL, NOSUB
• NPDTYPE, INTERPTYPE
• FORMAT (up to 20 characters)

Additional TABLE records

• FRML — LFORMAT
• FRMR — RFORMAT
• FETA — FIXEDETAS number list

Initial SCATTERPLOT record

(SCAT) (A4,4X,18I4)

Fields:

1. Scatterplot Step: conditional (0), unconditional (1), not implemented (2)
2. Number of families (1-20)

Individual SCATTERPLOT record

(SCAT) (A4,4X,9I8/8X,9I8)

Fields specify:

1. Index of abscissa data item (1-23)
2. Index of ordinate data item (1-23)
3. Partitioning: none (0), one-way (1), two-way (2)
4. Index of first separator
5. Index of second separator (for two-way)
6. Unit slope line flag (0 or 1)
7. FROM: first data record for scatter
8. TO: last data record for scatter
9. ORD0 flag: default (0), line through zero (1), no line (-1)
10. ABS0 flag: default (0), line through zero (1), no line (-1)
11. FIRSTONLY flag (0 or 1)
12. OBSONLY flag — only MDV=0 records (0 or 1)
13. NOSUB flag (-1, 0, 1)