model.create {RSiena} | R Documentation |
Creates a basic object which can be used as an argument for Siena07
model.create(fn = simstats0c, usesimstats0c=TRUE, projname = "Siena", MaxDegree = 0, useStdInits = FALSE, n3 = 1000, nsub = 4, maxlike = FALSE, diag = TRUE, condvarno = 0, condname = "", firstg = 0.2, cond = FALSE, findiff = FALSE, seed = NULL)
fn |
function to do one simulation in the Robbins-Monro algorithm |
usesimstats0c |
Boolean. If true the standard algorithm is being used which can be used with multiple processors. Just used for validation. |
projname |
character string name of project. No embedded spaces |
MaxDegree |
Named vector of maximum degree values for corresponding networks |
useStdInits |
Boolean. If TRUE, the initial values in the effects object will be ignored and default values used instead. |
n3 |
Number of iterations in phase 3. |
nsub |
Number of subphases in phase 2. |
maxlike |
whether to use maximum likelihood method or straightforward simulation |
diag |
Boolean: if FALSE, use the complete estimated derivative matrix in |
condvarno |
If conditional, the sequential number of the network or behavior variable on which to condition |
condname |
If conditional, the name of the dependent variable on
which to condition. Use one or other of condname or
condvarno to specify the variable. |
firstg |
initial value of scaling parameter in the Robbins-Monro procedure. |
cond |
Boolean. If TRUE, use conditional simulation |
findiff |
Boolean: If TRUE, estimate derivatives using finite differences. If FALSE use scores. |
seed |
Integer. Starting value of random seed. Not used if parallel testing. |
Model specification is done via this object for Siena7. This function creates an object with the correct elements. Those not available as arguments can be changed manually where desired.
Returns a model object of class "sienaModel" containing:
projname |
String value of name of project |
useStdInits |
Boolean, see above |
checktime |
Boolean, set to TRUE, report time in the phases or not |
n3 |
number of iterations in Phase 3 |
firstg |
Initial value of the scale in the Robbins-Munro algorithm |
maxrat |
value used to control the maximum size of the jumps |
maxmaxrat |
value used to control the maximum size of the jumps |
FRAN |
Simulation fn |
maxlike |
is FRAN using maximum likelihood? |
cconditional |
is FRAN using conditional estimation? |
condvarno |
number of dependent variable on which to condition |
condname |
name of dependent variable on which to condition |
FinDiff.method |
Boolean, should we calculate derivatives using finite differences? |
nsub |
Number of subphases in phase 2 |
diag |
Boolean - use just the diagonal of the derivative matrix? |
ModelType |
always 1 |
MaxDegree |
named vector of maximum degree values, or NULL |
Ruth Ripley
See http://www.stats.ox.ac.uk/~snijders/siena/
mymodel <- model.create(findiff=TRUE, fn = simstats0c)