Finding fragments of census tracts that are not in a place (in this case, Chicago) when other parts of the tract are in that place. Working against Census Reporter SQL schema as in http://censusreporter.tumblr.com/post/55886690087/using-census-data-in-postgresql

count_impacted_tracts output

 state | county | tract  | count 
-------+--------+--------+-------
 17    | 031    | 010400 |     1
 17    | 031    | 020702 |     1
 17    | 031    | 030101 |     1
 17    | 031    | 030102 |     1
 17    | 031    | 030103 |     1
 17    | 031    | 030104 |     1
 17    | 031    | 030601 |     1
 17    | 031    | 030604 |     1
 17    | 031    | 030706 |     1
 17    | 031    | 031300 |     1
 17    | 031    | 031400 |     1
 17    | 031    | 060800 |     1
 17    | 031    | 060900 |     1
 17    | 031    | 061902 |     1
 17    | 031    | 063200 |     1
 17    | 031    | 070101 |     1
 17    | 031    | 071400 |     1
 17    | 031    | 071500 |     1
 17    | 031    | 081202 |     1
 17    | 031    | 081403 |     1
 17    | 031    | 090200 |     1
 17    | 031    | 090300 |     1
 17    | 031    | 320600 |     1
 17    | 031    | 390700 |     1
 17    | 031    | 410100 |     1
 17    | 031    | 410900 |     1
 17    | 031    | 411000 |     1
 17    | 031    | 420100 |     1
 17    | 031    | 430102 |     1
 17    | 031    | 430700 |     1
 17    | 031    | 431400 |     1
 17    | 031    | 740400 |     1
 17    | 031    | 770500 |     4
 17    | 031    | 770602 |     3
 17    | 031    | 770700 |     3
 17    | 031    | 770800 |     2
 17    | 031    | 770902 |     2
 17    | 031    | 805600 |     1
 17    | 031    | 807900 |     1
 17    | 031    | 808100 |     1
 17    | 031    | 810400 |     2
 17    | 031    | 821402 |     2
 17    | 031    | 823304 |     3
 17    | 031    | 843900 |     1
 17    | 031    | 980000 |     7
 17    | 043    | 840000 |     2
 17    | 043    | 840801 |     1

count_impacted_tracts.sql

select g.state, g.county, g.tract, count(*)
from acs2011_5yr.geoheader g, 
	(select distinct state, county, tract from acs2011_5yr.geoheader where sumlevel = 80 and state = '17' and place = '14000') g2
where g.state = g2.state
and g.county = g2.county
and g.tract = g2.tract
and g.sumlevel = 80
and g.place != '14000'
group by g.state, g.county, g.tract
order by g.state, g.county, g.tract
;

identify_all_logrecnos output

 logrecno | state | county | tract  | place 
----------+-------+--------+--------+-------
     6461 | 17    | 031    | 010400 | 99999
     7784 | 17    | 031    | 020702 | 43744
     6462 | 17    | 031    | 030101 | 99999
     6463 | 17    | 031    | 030102 | 99999
     6464 | 17    | 031    | 030103 | 99999
     6465 | 17    | 031    | 030104 | 99999
     6466 | 17    | 031    | 030601 | 99999
     6467 | 17    | 031    | 030604 | 99999
     6468 | 17    | 031    | 030706 | 99999
     6469 | 17    | 031    | 031300 | 99999
     6470 | 17    | 031    | 031400 | 99999
     6471 | 17    | 031    | 060800 | 99999
     6472 | 17    | 031    | 060900 | 99999
     6473 | 17    | 031    | 061902 | 99999
     6474 | 17    | 031    | 063200 | 99999
     6475 | 17    | 031    | 070101 | 99999
     6476 | 17    | 031    | 071400 | 99999
     6477 | 17    | 031    | 071500 | 99999
     6478 | 17    | 031    | 081202 | 99999
     6479 | 17    | 031    | 081403 | 99999
     7726 | 17    | 031    | 090200 | 57875
     7727 | 17    | 031    | 090300 | 57875
     6480 | 17    | 031    | 320600 | 99999
     6481 | 17    | 031    | 390700 | 99999
     6482 | 17    | 031    | 410100 | 99999
     6483 | 17    | 031    | 410900 | 99999
     6484 | 17    | 031    | 411000 | 99999
     6485 | 17    | 031    | 420100 | 99999
     6486 | 17    | 031    | 430102 | 99999
     6487 | 17    | 031    | 430700 | 99999
     6488 | 17    | 031    | 431400 | 99999
     8369 | 17    | 031    | 740400 | 54820
     7480 | 17    | 031    | 770500 | 19642
     7490 | 17    | 031    | 770500 | 23256
     7495 | 17    | 031    | 770500 | 51089
     7516 | 17    | 031    | 770500 | 99999
     7697 | 17    | 031    | 770602 | 19642
     7739 | 17    | 031    | 770602 | 65819
     7742 | 17    | 031    | 770602 | 99999
     7740 | 17    | 031    | 770700 | 65819
     7623 | 17    | 031    | 770700 | 65819
     7627 | 17    | 031    | 770700 | 99999
     7624 | 17    | 031    | 770800 | 68081
     7628 | 17    | 031    | 770800 | 99999
     7856 | 17    | 031    | 770902 | 53377
     7615 | 17    | 031    | 770902 | 53377
     7730 | 17    | 031    | 805600 | 57875
     7785 | 17    | 031    | 807900 | 43744
     7793 | 17    | 031    | 808100 | 53000
     7859 | 17    | 031    | 810400 | 57875
     7860 | 17    | 031    | 810400 | 99999
     6647 | 17    | 031    | 821402 | 10513
     6649 | 17    | 031    | 821402 | 99999
     8348 | 17    | 031    | 823304 | 01010
     8368 | 17    | 031    | 823304 | 48554
     8398 | 17    | 031    | 823304 | 99999
     6496 | 17    | 031    | 843900 | 99999
     7710 | 17    | 031    | 980000 | 19642
     7485 | 17    | 031    | 980000 | 19642
     7494 | 17    | 031    | 980000 | 23256
     7741 | 17    | 031    | 980000 | 65819
     7626 | 17    | 031    | 980000 | 68081
     7522 | 17    | 031    | 980000 | 99999
     7756 | 17    | 031    | 980000 | 99999
     8637 | 17    | 043    | 840000 | 05248
     8644 | 17    | 043    | 840000 | 23256
     8642 | 17    | 043    | 840801 | 05248

identify_all_logrecnos.sql

select g.logrecno,g.state, g.county, g.tract, g.place 
from acs2011_5yr.geoheader g, 
	(select distinct state, county, tract 
	 from acs2011_5yr.geoheader 
	 where sumlevel = 80 and state = '17' and place = '14000') g2
where g.state = g2.state
and g.county = g2.county
and g.tract = g2.tract
and g.sumlevel = 80
and g.place != '14000'
order by g.state, g.county, g.tract, g.place
;