in_memory workspace bug

Discussion created by ehm119 on Apr 4, 2013
Latest reply on Apr 4, 2013 by ehm119
I posted this comment here but wanted to start a new thread, since this seems like a clear bug & I'd be interested if anyone else has come across it or found a solution.

I have a script that ranks the values in a field and writes the ranks (e.g largest = 1, smallest = n or vice versa) to another field.  When I do this on a FC stored in a fileGDB I get accurate results.  However, if I switch to the in-memory workspace, the rank values are sequential with the OBJECTIDs.  Can anyone replicate this, have similar problems, or know a workaround?

import arcpy
import sys

#Function to rank the values of a field write the rank to an existing field.
#Arguments include workspace path (workspace), FeatureClass name (FeatureClass), field with values to rank (valueField),
#and field to write values to (RankField).
#Order can be either ascending "A" (smallest value rank=1) or descending "D" (largest value rank =1)
class Rank:
 def __init__(self, workspace, FeatureClass, ValueField, RankField, Order):
  self.workspace = workspace
  self.fc = FeatureClass
  self.valueField = ValueField
  self.rankField = RankField
  self.Order = Order

 def rank(self):
   arcpy.env.workspace = self.workspace
   currRank = 0
   lastRank = 0
   lastVal = None

   #Since arcpy.da takes "ASC" and "DESC" instead of "A" and "D", reassign values if needed
   if self.Order == "A" or self.Order == "ASC":
    self.Order = "ASC"
   elif self.Order =="D" or self.Order == "DESC":
    self.Order = "DESC"
    arcpy.AddError('Sort order was not recognized.  Use "A" for ascending or "D" for descending.')
    print('Sort order was not recognized.  Use "A" for ascending or "D" for descending.')

   #arcpy.da version of rank script
   fields = (self.valueField, self.rankField)

   #sql_clause is a named parameter (don't need to add None values for other optional params.  ORDER BY sorts the ranking
   with arcpy.da.UpdateCursor(self.fc, fields, sql_clause=(None, 'ORDER BY {} {}'.format(self.valueField, self.Order))) as rows:
    for row in rows:
     currVal = row[0]
     if currVal != lastVal:                  #if the current value does not equal the previous value
      currRank += 1
      row[1] = currRank
     elif currVal == lastVal:
      row[1] = currRank
      print "Unexpected occurance"
     lastVal = currVal
     print "Finished rank script"

  except Exception as e:
   arcpy.AddError("The was a problem with the Rank Module... " + e.message)
   print("The was a problem with the Rank Module... " + e.message)