Bug fixes, enhancements for cor tool, it now uses the columnlist tool parameter.

tools/stats/cor.py

@@ -1,4 +1,5 @@
 #!/usr/bin/env python
+#Greg Von Kuster
 
 """
 Calculate correlations between numeric columns in a tab delim file.
@@ -7,7 +8,6 @@ usage: %prog infile output.txt columns method
 """
 
 import sys
-from Numeric import *
 from rpy import *
 
 def stop_err(msg):
@@ -77,9 +77,9 @@ def main():
         try:
             value = r.cor( array( matrix ), use="pairwise.complete.obs", method=method )
         except ValueError, exc:
-            stop_err("Computing correlation resulted in error: %s." %exc)
+            stop_err("Computing correlation resulted in error: %s." %str( exc ))
         except IndexError, exc:
-            stop_err("Computing correlation resulted in error: %s." %exc)
+            stop_err("Computing correlation resulted in error: %s." %str( exc ))
 
         for row in value:
             print >> out, "\t".join( map( str, row ) )
@@ -87,7 +87,10 @@ def main():
         out.close()
 
     if skipped_lines > 0:
-        print "Skipped %d invalid lines starting with line #%d.  Value '%s' in column %d is not numeric." % ( skipped_lines, first_invalid_line, invalid_value, invalid_column )
+        msg = "..Skipped %d lines starting with line #%d. " %( skipped_lines, first_invalid_line )
+        if invalid_value and invalid_column > 0:
+            msg += "Value '%s' in column %d is not numeric." % ( invalid_value, invalid_column )
+        print msg
 
 if __name__ == "__main__":
     main()

tools/stats/cor.xml

@@ -1,18 +1,14 @@
 <tool id="cor2" name="Correlation">
   <description>for numeric columns</description>
-  <command interpreter="python">cor.py $input1 $out_file1 $columns $method</command>
+  <command interpreter="python">cor.py $input1 $out_file1 $numeric_columns $method</command>
   <inputs>
-    <page>
-      <param format="tabular" name="input1" type="data" label="Dataset" help="Query missing? See TIP below"/>
-    </page>
-    <page>
-      <param name="columns" label="Numerical columns" type="select" multiple="True" dynamic_options="get_columns( input1 )" help="Multi-select list - hold the appropriate key while clicking to select multiple columns" />
-      <param name="method" type="select" label="Method">
-        <option value="pearson">Pearson</option>
-        <option value="kendall">Kendall rank</option>
-        <option value="spearman">Spearman rank</option>
-      </param>
-    </page>
+    <param format="tabular" name="input1" type="data" label="Dataset" refresh_on_change="True" help="Query missing? See TIP below"/>
+    <param name="numeric_columns" label="Numeric columns" type="columnlist" numerical="True" multiple="True" assoc_dataset="input1" help="Multi-select list - hold the appropriate key while clicking to select multiple columns" />
+    <param name="method" type="select" label="Method">
+      <option value="pearson">Pearson</option>
+      <option value="kendall">Kendall rank</option>
+      <option value="spearman">Spearman rank</option>
+    </param>
   </inputs>
   <outputs>
     <data format="txt" name="out_file1" />
@@ -23,7 +19,7 @@
     -->
     <test>
       <param name="input1" value="cor.tabular" />
-      <param name="columns" value="2,3" />
+      <param name="numeric_columns" value="2,3" />
       <param name="method" value="pearson" />
       <output name="out_file1" file="cor_out.txt" />
     </test>
@@ -99,5 +95,4 @@ This tool computes the matrix of correlation coefficients between numeric column
 
   So the correlation for our twenty cases is .73, which is a fairly strong positive relationship.
   </help>
-  <code file="cor_code.py"/>
 </tool>

tools/stats/cor_code.py

@@ -1,42 +0,0 @@
-
-def get_columns( input ):
-    columns = []
-    elems = []
-
-    for i, line in enumerate( file ( input.file_name ) ):
-        valid = True
-        if line and not line.startswith( '#' ): 
-            line = line.rstrip('\r\n')
-            elems = line.split( '\t' )
-    
-            """
-            Since this tool requires users to select only those columns
-            that contain numerical values, we'll restrict the column select
-            list appropriately.
-            """
-            if len(elems) > 0:
-                for col in range(1, input.metadata.columns+1):
-                    try:
-                        val = float(elems[col-1])
-                        valid = True
-                    except:
-                        val = elems[col-1]
-                        if val:
-                            if val.strip().lower() == "na":
-                                valid = True
-                            else:
-                                valid = False
-                        else:
-                            valid = False
-                    if valid:
-                        option = "c" + str(col)
-                        columns.append((option,str(col),False))
-            if len(columns) > 0:
-                """
-                We have our select list built, so we can break out of the outer most for loop
-                """
-                break 
-        if i == 30:
-            break # Hopefully we never get here...
-
-    return columns