def average_edit_distance(n):
import nltk.metrics.distance as d
edits = [ e["noteText"] for e in activity_logs_for_note(n) if e["action"] == "note-save"] ##//ActivityLog.objects.filter(noteid=n["id"],action="note-edit").order_by("when").values_list("noteText")
distances = []
if len(edits) > 1:
#print "edits: %s " % repr(edits)
for i in range(0,len(edits)-1):
if edits[i] is None or edits[i+1] is None:
continue
distances.append( d.edit_distance( edits[i], edits[i+1] ) )
if len(distances) > 0:
return make_feature("edit_distance",median(distances))
return make_feature('edit_distance',MISSING)
def average_edit_distance(n):
import nltk.metrics.distance as d
edits = [
e["noteText"] for e in activity_logs_for_note(n)
if e["action"] == "note-save"
] ##//ActivityLog.objects.filter(noteid=n["id"],action="note-edit").order_by("when").values_list("noteText")
distances = []
if len(edits) > 1:
#print "edits: %s " % repr(edits)
for i in range(0, len(edits) - 1):
if edits[i] is None or edits[i + 1] is None:
continue
distances.append(d.edit_distance(edits[i], edits[i + 1]))
if len(distances) > 0:
return make_feature("edit_distance", median(distances))
return make_feature('edit_distance', MISSING)
def s(counts):
print "min:%g max:%g mode:%g mean:%g median:%g var:%g" % (min(counts),max(counts),mode(counts),mean(counts),median(counts),var(counts))
return (min(counts),max(counts),mode(counts),mean(counts),median(counts),var(counts))
def s(counts):
print "min:%g max:%g mode:%g mean:%g median:%g var:%g" % (
min(counts), max(counts), mode(counts), mean(counts), median(counts),
var(counts))
return (min(counts), max(counts), mode(counts), mean(counts),
median(counts), var(counts))
