PROC TRAJ statement-options;

ID; VAR; INDEP;
MODEL; ORDER; START; WEIGHT;
RISK; REFGROUP;
TCOV; PLOTTCOV;
MIN; MAX; RORDER;    
/* cnorm options */
IORDER; EXPOS;       
/* zip options */
OUTCOME; OMODEL; OCOV;
/* outcome model */
DROPOUT; DCOV; OBSMAR;
/* dropout options */
MULTGROUPS; MULTRISK;
/* multi-trajectory */


statement-options follow PROC TRAJ in the first statement e.g.,

PROC TRAJ  DATA=ONE  OUT=OF  OUTPLOT=OP  OUTSTAT=OS  OUTEST=OE  ITDETAIL;

INPUT FILE: 

DATA= data for analysis, e.g. DATA=ONE.

OUTPUT FILES:

OUT= Group assignments and membership probabilities, e.g. OUT=OF.

OUTSTAT= Parameter estimates used by TRAJPLOT macro, e.g. OUTSTAT=OS.

OUTPLOT= Trajectory plot data, e.g. OUTPLOT=OP.

OUTEST= Parameter and covariance matrix estimates, e.g. OUTEST=OE.

OUTSTAT2=  (joint models) Parameter estimates used by the TRAJPLOT macro for the second model graph, e.g. OUTSTAT2=OS2.

OUTPLOT2= (joint models) Trajectory plot data for the second model graph, e.g. OUTPLOT2=OP2.

ADDITIONAL OPTIONS:

ITDETAIL displays minimization iterations for monitoring model fitting progress.

ALTSTART provides a second default start value strategy.


ID; Variables (typically containing information to identify observations) to place in the output (OUT=) data set, e.g. ID IDNO;

VAR; Dependent variables, measured at different times or ages (for example, hyperactivity score measured at age t,) e.g. VAR V1-V8;

INDEP; Independent variables (e.g. age, time) when the dependent (VAR) variables were measured, e.g. INDEP T1-T8;

MODEL; Dependent variable distribution (CNORM, ZIP, LOGIT) e.g. MODEL CNORM;

ORDER; Polynomial (0=intercept, 1=linear, 2=quadratic, 3=cubic) for each group, e.g. ORDER 2 2 2 0;

MIN; (CNORM) Minimum for censoring, e.g. MIN 1; If omitted, MIN defaults to zero.

MAX; (CNORM) Maximum for censoring, e.g. MAX 6; Required for CNORM

RORDER; (CNORM) Random growth curve parameters (-1=none, 0=intercept, 1=linear, 2=quadratic, 3=cubic) for each group.  For one set of random growth curve parameters, common to the groups, specify a single value. E.g. RORDER 0; or RORDER 0 0 0 0;

IORDER; (ZIP) Polynomial (0=intercept, 1=linear, 2=quadratic, 3=cubic) zero inflation probability logit for each group. For usual Poisson modeling, specify -1 for any group. For one inflation function common to the groups, specify one value in the IORDER statement. E.g. IORDER 1 1 1 1; or IORDER 2;

EXPOS; (ZIP) Exposure variables, e.g. EXPOS E1-E8;

START; Parameter starting values, e.g. START 1.10234  2.22309  1.34718    -.30128  .31274  1.12349   4.12311 .72918  -.36547  20 30 50;

RISK; Factors affecting the probability of group membership, e.g. RISK EDLEVEL;

REFGROUP; Controls the reference group (default = 1) when the risk statement is used, e.g. REFGROUP 2;

TCOV; Time varying covariates specified in sets, e.g. TCOV C1-C8; or TCOV C1-C8 D1-D8 E1-E8;

PLOTTCOV; Values for time varying covariates to calculate group trajectories. The number of values needs to be the same as the number of variables in the TCOV statement, e.g. PLOTTCOV 0 0 0 0 1 1 1 1; or PLOTTCOV 0 0 0 0 1 1 1 1  1 1 1 1 0 0 0 0  1 1 1 1 1 1 0 0;

WEIGHT; Weight variable for a weighted likelihood function. The covariance matrix is adjusted using a sandwich estimator. E.g. WEIGHT WGT;


The joint trajectory model uses the options shown above with a 2 suffix to specify the second model. See the joint trajectory model example.

MODEL2; VAR2; INDEP2; MIN2; MAX2; ORDER2; IORDER2; EXPOS2; RISK2; TCOV2; PLOTTCOV2;


MULTGROUPS; the number of multi-trajectory groups for the multi-trajectory model (2 to 6) e.g. MULTGROUPS 4;

MULTRISK; Factors affecting the probability of multi-trajectory group membership, e.g. MULTRISK FAMRISK;

Multi-Trajectory Model: the multi-trajectory model uses the options shown above with a 2 up to 6 suffix to specify the second up to sixth models