Update SVMLight script to also check file sanity.

MachLearnPort · May 23, 2015 · 0798e1c · 0798e1c
1 parent 1f6f4c2
commit 0798e1c
Showing 1 changed file with 8 additions and 0 deletions.
diff --git a/scripts/svmlight_sortcols.py → scripts/svmlight_check_and_sort.py b/scripts/svmlight_sortcols.py → scripts/svmlight_check_and_sort.py
@@ -26,6 +26,14 @@
             dsorted = sorted(d.iteritems(), key=itemgetter(0), reverse=False)
             with open(output, "a") as g:
                 g.write(target + ' ')
+                prev_i=-1
                 for (i,j) in dsorted:
+                    if(i == 0):
+                         print("Column index must be >= 1: " + dsorted)
+                         exit(1)
+                    if(i == prev_i):
+                         print("Repeats found: " + dsorted)
+                         exit(1)
+                    prev_i = i
                     g.write(str(i)+':'+str(j)+' ')
                 g.write('\n')