example2

The annotated example2 model specification file is included in the manual for convenience.

# This is a model specification file.  Please see the GENBLIS manual for
# details.  The manual is available at
# http://data.fas.harvard.edu/jsekhon/genblis/.  Your distribution of
# GENBLIS ought to have come with a postscript version of the manual.

# Comments begin with a "#" and continue until the end of the line.
# Comments are not allowed to occur between keywords and their arguments.

# The model specification file begins with the GENERAL SETUP SECTION.

# Required.  This keyword precedes the path to where the dataset is
# located.  There can be a maximum of 4096 characters in any path name used
# in GENBLIS.  There is no default path.

datafile example2.dat


# Required.  This variable must be set to the number of observed variables
# in the data.  The number is equal to the number of variables which must
# be read in from the data file.  There is no default.
  
observedvars 6


# Required.  The variable sets the number of latent variable present in the
# linear structure model to be estimated.  There is not default.

latentvars 1

# Optional.  If you don't want any bootstrap estimates, just uncomment the
# "noboots" keyword below.
# noboots

# Optional.  If you want neither bootstrap estimates nor GENBLIS to use the
# control.big file, just uncomment the "nobig" keyword below.
# nobig


# Optional. This option sets the number of generations after GENBLIS thinks
# it has converged it ought to keep going.  This number must be an integer.
# GENBLIS thinks it has converged if the gradients at the best solution
# found so far are below a criterion defined below.  It is often good to
# keep GENBLIS keeps working after this point because the evolutionary
# program (EP) portion of GENBLIS continues to provide useful information.
# Obviously, the higher this number is, the greater security one has that a
# global optimal has been found.  The default value is 5.

genconverge 1


# Optional.  This variable sets the maximum number of generations.  Recall
# that theory states that the size of the genetic population is of greater
# importance than then number of generations. The asymptotics are
# primarily in population size.  The default value is 100.

genmax 20


# Required.  The "model" keyword begins the Linear Structure Model Definition
# part of this file.  This section is required.

model

# The model which is implemented here is
#
# y1 = l1 * latent1 + e1,
# y2 = l2 * latent1 + e2,
# y3 = 1  * latent1 + e3,  y3 is used to identify the model and to scale
#                          the latent variable
# y4 = l3 * latent1 + e4,
# y5 = l4 * latent1 + e5,
# y6 = l5 * latent1 + e6,
#
# the variance of latent1 is psi1,
#
# where, y1 through y6 are the six observed variables, "latent1" is the one
# latent variable, and e1 through e6 are the random errors, and l1 through
# l5 are the coefficients which scale the latent variable to the
# corresponding observed variable.  e1 through e6 have variances which are
# estimated as the diagonal elements of te.  There are no estimated
# covariances.
#
# The bootstrap goodness-of-fits tests (presented in the example2.lst file)
# show that this model does NOT fit the data.


# Setup Lambda Y.  No comments are allowed until all of the arguments to
# the keyword "ly" have been entered.
# This defines a one factor model.  Lambda Y[3,1] is fixed to one for
# identification.  All other loadings are estimated.
ly
1 1 0
2 1 0
3 1 1
4 1 0
5 1 0
6 1 0

# Setup Theta Epsilon.  No comments are allowed until all of the arguments to
# the keyword "te" have been entered.  All of the diagonal elements of Theta
# Epsilon are estimated.
te
-1 -1 0

# Setup Psi.  No comments are allowed until all of the arguments to
# the keyword "psi" have been entered.  The random error of the one latent
# variable is estimated by Psi[1,1].
psi
1 1 0


# Required.  The keyword "bounds" begins the Parameter Bounds portion of
# the model specification file.  No comments are allowed until all of the
# bounds have been listed.  Format: parameter-number lower-bound
# upper-bound

bounds
1   -1.00      20.10 
2   -1.00      20.10 
3   -1.00      20.10 
4   -1.00      20.10 
5   -1.00      20.10
6   0.00      0.90 
7   0.00      0.90 
8   0.00      0.90 
9   0.00      0.90 
10  0.00      0.90 
11  0.00      0.90 
12  0.00      0.90

# Optional.  No Starting Values are provided.  If there had been a starting
# value section, it would have been started by the "starting_values"
# keyword


# Optional.  The keyword "control" begin the Genetic Operator Controls
# setup for the control file.  None of the following control files is
# required.  This section defines the four control file used by GENBLIS.
# GENBLIS has four different control files defined by the following
# keywords: control (used for the original sample); control.big (used in the
# original sample and when there is a convergence failure in either the
# jackknives or the bootstraps); control.jack (used for the jackknives);
# and control.boot (used for the bootstraps).  Each control defines 9
# different variables: the number used of each of 8 operators, and the
# gradient convergence tolerance.  The gradient convergence tolerance is a
# number which all of the gradients have to be below for GENBLIS to think that
# it has found a minimum.  The default tolerance is 0.00002.
#
# The syntax is "control-file-keyword" followed by up to 9 numbers.  The
# first 8 numbers define the number of operators of the chosen type.  The
# 9th number is the gradient tolerance.  Obviously, the first 8 numbers
# must be integers.  The first 8 numbers refer to the operators in the
# following order:
#
# uniform mutation (50)
# boundary mutation (0)
# non-uniform mutation (50)
# polytope crossover (50)
# multiple point simple crossover (50)
# whole non-uniform mutation (50)
# heuristic crossover (50)
# local-minimum crossover (0).
#
# The number in brackets is the default number of operators of each type
# that GENBLIS uses.  If the "control" file is set and another of the given
# files is not set, it is set equal to the control file.  If only part of a
# control file's arguments is listed (e.g., "control.big" is only followed
# by 3 numbers) the rest of the operators and the gradient tolerance are
# set to equal the default values listed above.


control
50      0       50      50      50      50      50      0

control.big
60      0       60      60      60      60      60      0

control.jack
2       0       2       2       2       2       20      0

control.boot
40      0       40      40      40      40      40      0