[rkward-cvs] SF.net SVN: rkward:[2626] trunk/rkward/rkward/plugins/analysis

tfry at users.sourceforge.net tfry at users.sourceforge.net
Tue Aug 25 13:43:35 UTC 2009


Revision: 2626
          http://rkward.svn.sourceforge.net/rkward/?rev=2626&view=rev
Author:   tfry
Date:     2009-08-25 13:43:35 +0000 (Tue, 25 Aug 2009)

Log Message:
-----------
Slight enhancements to documentation, labels, and in one case defaults.

Modified Paths:
--------------
    trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.rkh
    trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.xml
    trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.rkh
    trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.xml
    trunk/rkward/rkward/plugins/analysis/outliers/outlier.php
    trunk/rkward/rkward/plugins/analysis/outliers/outlier.rkh
    trunk/rkward/rkward/plugins/analysis/variances/levene_test.php

Modified: trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.rkh
===================================================================
--- trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.rkh	2009-08-25 13:41:37 UTC (rev 2625)
+++ trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.rkh	2009-08-25 13:43:35 UTC (rev 2626)
@@ -5,15 +5,15 @@
 	</summary>
 
 	<usage>
-		Chose the data to perform the Dixon test for outlier. You can chose multiple data. See <link href="rkward://rhelp/dixon.test"/>.
+		Chose the data to perform the Dixon test for outlier. You can chose multiple vectors. See <link href="rkward://rhelp/dixon.test"/>.
 	</usage>
 
 	<settings>
 		<caption id="tab_variables"/>
 		<setting id="x">Select the data to be computed. The vectors need to be numeric, and can be of different length but there is a limit.</setting>
 		<caption id="tab_options"/>
-		<setting id="type">This is specyfying the variant of test to be performed.</setting>
-		<setting id="two_sided">Here you can specify the alternative hypothesis. It must be one of "two-sided" or "not two-sided".</setting>
+		<setting id="type">Specifies the variant of test to be performed. See <link href="rkward://rhelp/dixon.test"/> for details. The default value (0) means that a suitable variant is selected automatically, according to the size of the sample.</setting>
+		<setting id="two_sided">Should the test be treated as two-sided (default)?</setting>
 		<setting id="opposite">Here you can define if you want to check not the value with largest difference from the mean, but opposite.</setting>
 		<setting id="descriptives">Print additional descriptive statistics (mean, sd, median, min, max) to help with interpretation of the test result (purely informational, does not affect the test). Missing values will be ignored.</setting>
 		<setting id="length">Show the total length of the data, and the number of missing values (purely informational, does not affect the test)?</setting>

Modified: trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.xml
===================================================================
--- trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.xml	2009-08-25 13:41:37 UTC (rev 2625)
+++ trunk/rkward/rkward/plugins/analysis/outliers/dixon_test.xml	2009-08-25 13:43:35 UTC (rev 2626)
@@ -19,7 +19,7 @@
 				<column>
 					<frame label="test specific settings">
 						<dropdown id="type" label="Variant of test">
-							<option value="0" label="0"/>
+							<option value="0" label="0 (Automatic)"/>
 							<option value="10" label="10 (sample size 3-7)"/>
 							<option value="11" label="11 (sample size 8-10)"/>
 							<option value="12" label="12"/>

Modified: trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.rkh
===================================================================
--- trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.rkh	2009-08-25 13:41:37 UTC (rev 2625)
+++ trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.rkh	2009-08-25 13:43:35 UTC (rev 2626)
@@ -10,10 +10,10 @@
 
 	<settings>
 		<caption id="tab_variables"/>
-		<setting id="x">Select the data to be computed. The vectors need to be numeric, and can be of different length but there is a limit.</setting>
+		<setting id="x">Select the data to be computed. The vectors need to be numeric, and may be of different length.</setting>
 		<caption id="tab_options"/>
-		<setting id="type">This is specyfying the variant of test to be performed. Please refere to the R-help here.</setting>
-		<setting id="two_sided">Here you can specify the alternative hypothesis. It must be one of "two-sided" or "not two-sided"</setting>
+		<setting id="type">Variant of test to be performed. One or two outliers can be detected, on the same, or on opposite tails.</setting>
+		<setting id="two_sided">Should the test be treated as two-sided?</setting>
 		<setting id="opposite">Here you can define if you want to check not the value with largest difference from the mean, but opposite.</setting>
 		<setting id="descriptives">Print additional descriptive statistics (mean, sd, median, min, max) to help with interpretation of the test result (purely informational, does not affect the test). Missing values will be ignored.</setting>
 		<setting id="length">Show the total length of the data, and the number of missing values (purely informational, does not affect the test)?</setting>

Modified: trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.xml
===================================================================
--- trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.xml	2009-08-25 13:41:37 UTC (rev 2625)
+++ trunk/rkward/rkward/plugins/analysis/outliers/grubbs_test.xml	2009-08-25 13:43:35 UTC (rev 2626)
@@ -21,10 +21,10 @@
 						<radio id="type" label="Variant of test">
 							<option value="10" label="10 (test for one outlier)"/>
 							<option value="11" label="11 (test for two outliers on opposite tails)"/>
-							<option value="20" label="20 (test for two outliers in one tail)"/>
+							<option value="20" label="20 (test for two outliers on one tail)"/>
 						</radio>
 						<checkbox id="opposite" label="Opposite" value_unchecked="FALSE" value="TRUE" checked="false" />
-						<checkbox id="two_sided" label="two-sided (default)" value_unchecked="FALSE" value="TRUE" checked="true" />
+						<checkbox id="two_sided" label="two-sided" value_unchecked="FALSE" value="TRUE" checked="false" />
 					</frame>
 					<frame label="additional settings">
 						<checkbox id="descriptives" label="Show additional descriptives" value_unchecked="0" value="1" checked="false" />

Modified: trunk/rkward/rkward/plugins/analysis/outliers/outlier.php
===================================================================
--- trunk/rkward/rkward/plugins/analysis/outliers/outlier.php	2009-08-25 13:41:37 UTC (rev 2625)
+++ trunk/rkward/rkward/plugins/analysis/outliers/outlier.php	2009-08-25 13:43:35 UTC (rev 2626)
@@ -43,7 +43,7 @@
 
 function printout () {
 ?>
-rk.header ("Outlier",
+rk.header ("Find potential outlier",
 	parameters=list ("Opposite", "<? getRK ("opposite"); ?>"))
 rk.results (results)
 <?

Modified: trunk/rkward/rkward/plugins/analysis/outliers/outlier.rkh
===================================================================
--- trunk/rkward/rkward/plugins/analysis/outliers/outlier.rkh	2009-08-25 13:41:37 UTC (rev 2625)
+++ trunk/rkward/rkward/plugins/analysis/outliers/outlier.rkh	2009-08-25 13:43:35 UTC (rev 2626)
@@ -1,18 +1,18 @@
 <!DOCTYPE rkhelp>
 <document>
 	<summary>
-		This test finds Outlier.
+		Shows the value with largest difference from sample mean (potential outlier).
 	</summary>
 
 	<usage>
-		Chose the data to find the value with the largest difference from the sample mean. You can test multiple data sets. See <link href="rkward://rhelp/outlier"/>.
+		Chose the data to find the value with the largest difference from the sample mean. You can select multiple vectors. See <link href="rkward://rhelp/outlier"/>.
 	</usage>
 
 	<settings>
 		<caption id="tab_variables"/>
-		<setting id="x">Select the data to be computed. The vectors need to be numeric and can have missing values included. For the calculation missing values are removed from the data.</setting>
+		<setting id="x">Select the data to be computed. The vectors need to be numeric. Any missing values are ignored in the calculation.</setting>
 		<caption id="tab_options"/>
-		<setting id="opposite">Set to give the opposite value. If the largest value has maximum difference from the mean, it gives smallest and vice versa.</setting>
+		<setting id="opposite">Select this option to give the opposite value. If the largest value has maximum difference from the mean, it gives smallest and vice versa.</setting>
 		<setting id="descriptives">Print additional descriptive statistics (mean, sd, median, min, max) to help with interpretation of the test result (purely informational, does not affect the test). Missing values will be ignored.</setting>
 		<setting id="length">Show the total length of the data, and the number of missing values (purely informational, does not affect the test)?</setting>
 	</settings>

Modified: trunk/rkward/rkward/plugins/analysis/variances/levene_test.php
===================================================================
--- trunk/rkward/rkward/plugins/analysis/variances/levene_test.php	2009-08-25 13:41:37 UTC (rev 2625)
+++ trunk/rkward/rkward/plugins/analysis/variances/levene_test.php	2009-08-25 13:43:35 UTC (rev 2626)
@@ -14,7 +14,7 @@
 ?>
 names <- rk.get.description (<? getRK ("y"); ?>, <? getRK ("group"); ?>)
 
-rk.header ("Levene's Test",names)
+rk.header ("Levene's Test", list ("response variable", names[1], "groups", names[2]))
 
 rk.print (result)
 <?


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




More information about the rkward-tracker mailing list