import dataframe.*
If the jar files are in a different folder/directory, then use the full path, e.g.
| javaaddpath 'c:\SSS\sss.jar' |
and so forth.
Running SSS from command line in Matlab
Many users will setup an input text file -- say, setup.txt -- either directly or with the GUI and then run
the SSS model search from within Matlab via
| !modelsearch.exe setup.txt |
Using the GUI
- start the GUI with commands
| mySSS=sss.Model; mySSS.start |
- use the GUI to customise input/ouput parameter values in setup file.
- Save and Load to create or input setup text files.
- hit the Run button in the GUI to start the SSS model search.
- hit the Stop button in the GUI if you want to abandon a running SSS model search.
Setup file of input parameters
The parameter setup/input file (named setup.txt above) is a flat text file with each line representing a parameter
(name, value) pair for a predefined set of parameters. The order of
parameters in the file is not important. User can also comment out a line by
adding # at the beginning. Each line is in the format
ParameterName = Value
where ParameterName is one of the names described bellow and
Value takes a string or numberic values depending on the nature of given
parameter. When a path is used as a parameter value, the spaces in the path
will be ignored. SSS will NOT
work with a path that has spaces in it.
| ----------------- |
| Inputs: |
NOBSERVATIONS | n = total sample size |
| NVARIABLES | p = total number of predictor variables |
| DATAFILE | tab delimited predictor data (n rows, p columns) |
| RESPONSEFILE | response variable (n values, tab delimited row or single column) |
| WEIGHTSFILE | weight vector (n values of 0/1; 1 indicates samples to be used in model fit) |
| CENSORFILE | 0/1 indicators of right-censoring (0) versus observed (1) in case of survival data
|
| ----------------- |
| Output: |
OUTFILE | list of selected best models |
| ITEROUT | details of models visited at each SSS iteration |
| SUMMARYFILE | summary of models and parameter estimation |
| DEBUGOUT | output debug information to standard output, or not |
| | |
| | |
| | |
| | |
| ----------------- |
| Model/Search: |
MODTYPE | Model Type: 1=linear, 2=binary/logistic, 3=Weibull survival |
| PSTART | Initial Model Size: number of predictors for model to start SSS |
| PMAX | Maximum Model Size: maximum number of predictors in any model |
| PRIORMEANP | Prior mean of number of included variables: this is the key sparsity control
parameter (PRIORMEANP=v means that each variable is "in the model" with prior probability v/p) |
| NBEST | Number of Best Models to be saved/recorded |
| ITERS | Total Number of SSS iterations |
| ONEVAR | indicator to include all 1-variable models in search |
| ----------------- |
| Annealing*: |
(replace, innerAnneal1) | annealing parameter for variable replacement |
| (delete, innerAnneal2) | annealing parameter for variable deletion |
| (add, innerAnneal3) | annealing parameter for variable addition |
| (outer, outerAnneal) | annealing parameter for second level model selection |
| | |
| | |
| ----------------- |
*Annealing parameters can generally be left at the suggested defaults in the example
files here. See the paper for additional discussion.
Output files
The key output file is the model summary file -- a flat txt file summarising the posterior
distributions within and across the models summarised. See the examples in the
Matlab examples file examples.m and the three summary support .m script files for these
examples. The SSS search explores linear models with standardised y and x data and so the output summaries relate to the
standardised models with no intercept. In contrast the binary and survival models include intercepts.
The model summary text file output has the following information.
Each row is one model of the "top models" ordered in decreasing order of posterior probability. The number of columns is defined
by the largest model, and entries are NA/NaN for models smaller than the largest
In each model/row the entries are as follows:
Linear regression models:
- element 1 - dimension of the model = number of predictors p for this model
- element 2 - log posterior probability of this model (the "score")
- elements 3:(2+p) - the indices of the p variables in this model
- elements (3+p):(2+2p) - posterior mean of the regression parameter vector beta (no intercept)
- elements (3+2p):(2+2p+p*p) - posterior variance matrix of beta in vectorised form (no intercept)
- final two elements: - (s,d), the residual SD estimate, and the posterior degees of freedom
Binary logistic regression models:
- element 1 - dimension of the model = number of predictors p for this model
- element 2 - log posterior probability of this model (the "score")
- elements 3:(2+p) - the indices of the p variables in this model
- elements (3+p):(3+2p) - posterior mode of the regression parameter vector beta (includes intercept)
- elements (4+2p):(4+4p+p*p) - estimated posterior variance matrix of beta in vectorised form (includes intercept)
Survival (Weibull) regression models:
- element 1 - dimension of the model = number of predictors p for this model
- element 2 - log posterior probability of this model (the "score")
- elements 3:(2+p) - the indices of the p variables in this model
- element 3+p: - the posterior mean of the Weibull index parameter alpha in this model
- elements (4+p):(4+2p) - posterior mode of the regression parameter vector beta (includes intercept)
- elements (5+2p):(8+6p+p*p) - estimated posterior variance matrix of (alpha,beta) (beta includes intercept) in vectorised form
|