[rkward-cvs] SF.net SVN: rkward:[2719] trunk/rkward/tests/item_response_theory

[m-eik at users.sourceforge.net]

Revision: 2719
          http://rkward.svn.sourceforge.net/rkward/?rev=2719&view=rev
Author:   m-eik
Date:     2009-11-30 20:59:34 +0000 (Mon, 30 Nov 2009)

Log Message:
-----------
irt: ...as goes for the gpcm tests standards

Added Paths:
-----------
    trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkcommands.R
    trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkout

Added: trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkcommands.R
===================================================================
--- trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkcommands.R	                        (rev 0)
+++ trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkcommands.R	2009-11-30 20:59:34 UTC (rev 2719)
@@ -0,0 +1,16 @@
+local({
+## Prepare
+  require(ltm)
+## Compute
+estimates.gpcm <- gpcm(subset(Science, select=c("Work", "Industry", "Future", "Benefit")), constraint="rasch")
+## Print result
+rk.header ("GPCM parameter estimation")
+rk.print (paste("Call: <code>",deparse(estimates.gpcm$call, width.cutoff=500),"</code>"))
+rk.print ("<h4>Coefficients:</h4>")
+rk.print (coef(estimates.gpcm))
+rk.print (paste("Log-likelihood value at convergence:",round(estimates.gpcm$log.Lik, digits=1)))
+# keep results in current workspace
+estimates.gpcm <<- estimates.gpcm
+})
+.rk.rerun.plugin.link(plugin="rkward::par_est_gpcm", settings="chk_save.state=save\nchk_select.state=select\nconstraint.string=rasch\nepshess.real=0.000001\nghk_gpcm.real=21.00\ninp_items.available=Science[[\\\"Work\\\"]]\\nScience[[\\\"Industry\\\"]]\\nScience[[\\\"Future\\\"]]\\nScience[[\\\"Benefit\\\"]]\nirtparam.state=TRUE\niterqn_gpcm.real=150.00\nnaaction.state=\nnumrderiv.string=fd\noptimeth.string=BFGS\noptimizer.string=optim\nsave_name.selection=estimates.gpcm\nstartval.string=NULL\nverbose.state=\nx.available=Science", label="Run again")
+.rk.make.hr()

Added: trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkout
===================================================================
--- trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkout	                        (rev 0)
+++ trunk/rkward/tests/item_response_theory/RKTestStandard.GPCM_parameter_estimation.rkout	2009-11-30 20:59:34 UTC (rev 2719)
@@ -0,0 +1,17 @@
+<h1>GPCM parameter estimation</h1>
+DATE<br>
+<p class='character'>Call: <code> gpcm(data = subset(Science, select = c("Work", "Industry", "Future", "Benefit")), constraint = "rasch") </code></p><p class='character'><h4>Coefficients:</h4></p>
+<p align= center >
+<table cellspacing=0 border=1><caption align=bottom class=captiondataframe></caption><tr><td>
+	<table border=0 class=dataframe>
+	<tbody> <tr class= firstline > <th></th><th>Catgr.1</th><th>Catgr.2</th><th>Catgr.3</th><th>Dscrmn</th> </tr>
+ <tr><td class=firstcolumn>Work</td><td class=cellinside>-1.80</td><td class=cellinside>-0.90</td><td class=cellinside> 1.80</td><td class=cellinside> 1.00</td></tr>
+ <tr><td class=firstcolumn>Industry</td><td class=cellinside>-2.54</td><td class=cellinside>-1.76</td><td class=cellinside> 0.20</td><td class=cellinside> 1.00</td></tr>
+ <tr><td class=firstcolumn>Future</td><td class=cellinside>-2.50</td><td class=cellinside>-1.39</td><td class=cellinside> 1.09</td><td class=cellinside> 1.00</td></tr>
+ <tr><td class=firstcolumn>Benefit</td><td class=cellinside>-2.33</td><td class=cellinside>-0.88</td><td class=cellinside> 1.31</td><td class=cellinside> 1.00</td></tr>
+ 
+	</tbody>
+</table>
+ </td></table>
+ <br>
+<p class='character'>Log-likelihood value at convergence: -1756.8</p>
\ No newline at end of file


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.