Dissolve/ based on condition

[HushamMohamed]

Hello, 

I would like to dissolve large  polygon feature data set(25000 units) based on 2 conditions:

1- by attribute value partially matched,  (UNIT field)  

for example

id	Unit
1	CLAY BROWN UNIT 2
2	CLAY BROWN UNIT 3
3	CLAY BROWN UNIT 4 sub unit 6
4	CLAY BROWN UNIT 5

2- only adjacent polygon should be dissolved. ( warning topology is not clean 100% , small gaps may appear)

example only rows 1,2,3 are adjacent polygon.

I would appreciate any suggestion. 

Thank you

[DavidPike]

Hi,

The dissolve will only dissolve adjacent features so condition 2 will be met anyway.

Regards gaps I would look at the integrate tool but be warned this will alter the existing data. I highly recommend making a copy before u do anything with it. It will also probably destroy and topology u have set up.

If u mean a partial field match I would probably add a new field such as higher level soil unit etc. And run a field calculator with a python code block to get the matches. I can write this when I get some time

[HushamMohamed]

Thank for your your helpful information, 

The dissolve will only dissolve adjacent features so condition 2 will be met anyway.

Yes  I mean partial field match, from the example table (Clay Brown Unit) is partial match. another Example 

1- East Soil corrected 6th addition

2- East Soil corrected 5th addition

3- East Soil corrected 5th sub plat

only 1,2 should be dissolved.

 I would love to get that python script, to test,  

Thank you.

[HushamMohamed]

defining multiple flags  in code block is a challenge, do I need to use regex. 

[DavidPike]

How many entries are there to match to? If you list them I'll put them in the code to match. If theres hundreds I'll think of a better solution

[HushamMohamed]

Thanks,  I am afraid they are  hundreds. I have 25000 records, (10 feature classes).

[DavidPike]

Is there an identifiable pattern to break/group the string by? If not I cant really see a solution.

[HushamMohamed]

Yes that is the problem, there is no identifiable pattern, 

Thank you very much.

[DanPatterson_Retired]

If there is no pattern, you may have to break them into groups and process accordingly.  Sometimes there is no one-solves-all.  You indicate that you have 10 featureclasses but 25000 records.  It appears you are looking for a substring within a string.  If it were you could use that information to parse the data first prior to determining whether to dissolve.

cases = ['East Soil corrected 6th addition',
         'East Soil corrected 5th addition',
         'East Soil corrected 5th sub plat']

for c in cases:
    if 'East Soil corrected' in c and 'addition' in c:
        print("dissolve")
    else:
        print("don't dissolve")
        
dissolve
dissolve
don't dissolve‍‍‍‍‍‍‍‍‍‍‍‍‍

[HushamMohamed]

Thank you, using Case statement is not an option,  keeping up with it for anything new may be a pain.