Resumen
Creates a geoprocessing object. The extra arguments can be used to specify additional requirements for the object creation such as the number of columns in the ValueTable object.
Debate
Nota:
Instead of using CreateObject, it is simpler and more direct to use the equivalent ArcPy class. For instance, instead of arcpy.CreateObject("array"), use arcpy.Array().
Sintaxis
CreateObject (name, {options})| Parámetro | Explicación | Tipo de datos | 
| name | The name of the object to be created (ArcSDESQLExecute, Array, Extent, FeatureSet, Field, FieldInfo, FieldMap, FieldMappings, Geometry, NetCDFFileProperties, Point, RecordSet, Result, SpatialReference, and ValueTable). | String | 
| options | Optional arguments depending on the object being created. | Object | 
| Tipo de datos | Explicación | 
| Object | The object returned depends on type of object specified in the first parameter. | 
Muestra de código
Use value table to hold feature class names and ranks for the Union tool.
import arcpy
# Set the workspace. List all of the feature classes in the dataset.
arcpy.env.workspace = "c:/data/landbase.gdb/wetlands"
fcs = arcpy.ListFeatureClasses()
# Create the value table for the Analysis Union tool with 2 columns.
vtab = arcpy.CreateObject("valuetable", 2)
# Iterate through the list of feature classes.
for fc in fcs:
    # Update the value table with a rank of 2 for each record, except
    #   for BigBog
    if fc.lower() != "bigbog":
        vtab.addRow(fc + " 2")
    else:
        vtab.addRow(fc + " 1")
# Union the wetlands feature classes with the land use feature class
# to create a single feature class with all of the wetlands and land
# use data.
vtab.addRow("c:/data/landbase.gdb/land_use 2")
arcpy.analysisUnion(vtab, "c:/data/landbase.gdb/wetlands_use")