List unused domain values in a domain via Python

[Brian_McLeer]

Any ideas on how to list unused domain values from within a coded domain?

An example is we had PKWY EAST in a domain for street suffixes. I would like to generate a list of domain values that are not being used in a feature class. 

[deleted-user-qpvAI3Fo0MKR]

It's outside the scope of your request, but identifying orphaned domains is out of the box functionality for Laurel Hills GIS GenSentry application.

[forestknutsen1]

What about comparing all used domain values with a table of all domain values?

• Get a table of all domains  Table To Domain—Help | ArcGIS for Desktop 
• Get a unique list of all domains with Summarize Attributes—GeoAnalytics Tools | ArcGIS Desktop, you may have to merge and re-summarize if your domain is on more then one table
  • Or better yet if you have a sde you can use sql select distinct, etc. with cx_oracle  Welcome to cx_Oracle’s documentation! — cx_Oracle 7.0.0-dev documentation 
• Left join the list of all domains and the summarized (select distinct) and then select where null

[deleted-user-qpvAI3Fo0MKR]

Clever 🙂

[RandyBurton]

Similar to Forest's approach, I would load the feature's field with the domains you are interested in into a dictionary using code such as:

import arcpy

fc = <your feature>
field = <domain field>

d = {}

with arcpy.da.SearchCursor(fc, (field)) as rows:
    for row in rows:
        if row[0] not in d:
            d[row[0]] = 1
        else:
            d[row[0]] += 1

for k, v in d.iteritems():
    # something like if k not in domainTable ....
    # or if domainTable[value] not in d.keys() ...‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

And then you can compare with the table to domain information.  This would also count the number of rows with specific values.

[RandyBurton]

And in place of TableToDomain/DomainToTable, see ListDomains on how to get a list of domains and coded/max/min values.  You can use something like ListFields to get the domain name a field uses (field.domain).

[RandyBurton]

Here's a sample script using ideas in my previous posts:

import arcpy

gdb = r"C:\Path\To\file.gdb"
arcpy.env.workspace = gdb # set environment for arcpy

# gdb domains to dictionary # # # # # # #
domDict = {} # empty dictionary
domains = arcpy.da.ListDomains(gdb)
for domain in domains:
    if domain.domainType == 'CodedValue':
        if domain.name not in domDict:
            vList = [] # empty list
            coded_values = domain.codedValues
            for val, desc in coded_values.items():
                vList.append({val:desc})
        domDict[domain.name] = vList
# print domDict

# read feature's fields and domains information # # # # # # #
fc = 'MyFeature' # a feature in the geodatabase
fields = arcpy.ListFields(fc)
fldDict = {} # empty dictionary
for field in fields:
    if len(field.domain):
        fldDict[field.name] = field.domain
# print fldDict

# count values in fields with domains # # # # # # #
domCount = {}
fldList = list(fldDict.keys())
with arcpy.da.SearchCursor(fc, fldList) as rows:
    for row in rows:
        for i, f in enumerate(fldList):
            # print i, fldList, row # index, field name, field value
            if fldList[i] not in domCount:
                domCount[fldList[i]] = {}
            if row[i] not in domCount[fldList[i]]:
                domCount[fldList[i]][row[i]] = 1
            else:
                domCount[fldList[i]][row[i]] += 1
del rows # release any locks
# print domCount

# output the results # # # # # # #
for k, v in domCount.iteritems():
    print "Field: {}".format(k)
    print "  Domain: {}".format(fldDict[k])
    for dLst in domDict[fldDict[k]]:
        for k1, v1 in dLst.iteritems():
            if k1 in v:
                print "    Key:  {} - Description: {} - Count: {}".format(k1, v1, v[k1])
            else:
                print "    Key:  {} - Description: {} - Count: {}".format(k1, v1, 0)

# may need to add code to check for invalid codes in fields ?‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

And my test results:

Field: MyType
  Domain: MyCode
    Key:  1234 - Description: One - Count: 3
    Key:  2345 - Description: Two - Count: 2
    Key:  3456 - Description: Three - Count: 0
    Key:  6789 - Description: Six - Count: 0
    Key:  4567 - Description: Four - Count: 0
    Key:  5678 - Description: Five - Count: 0
    Key:  7890 - Description: Seven - Count: 1
Field: MyColor
  Domain: MyColor
    Key:  Y - Description: Yellow - Count: 1
    Key:  P - Description: Purple - Count: 0
    Key:  B - Description: Blue - Count: 1
    Key:  R - Description: Red - Count: 2
    Key:  G - Description: Green - Count: 2
‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍