TIGER Geocoder data loading misc

[darkblueb]

svn rev 13095 PostgreSQL 9.4b4 Ubuntu 14.04 — (see attachment)

[darkblueb]

load California data, misc notes

[darkblueb]

create extension postgis_tiger_geocoder;

executes, but address_standardizer is not added.

addr_test_94=# \df add* => no results

executing create extension address_standardizer;

addr_test_94=# create extension address_standardizer ;
CREATE EXTENSION

##--
addr_test_94=# select * FROM parse_address('1 Devonshire Place, Boston, MA 02109-1234');
 num |      street      | street2 |      address1      |  city  | state |  zip  | zipplus | country 
-----+------------------+---------+--------------------+--------+-------+-------+---------+---------
 1   | Devonshire Place |         | 1 Devonshire Place | Boston | MA    | 02109 | 1234    | US
(1 row)

##--
addr_test_94=# SELECT *  FROM standardize_address('tiger.pagc_lex', 
   'tiger.pagc_gaz', 'tiger.pagc_rules', 'One Devonshire Place, PH 301, Boston, MA 02109-1234');
   
 building | house_num | predir | qual | pretype |    name    | suftype | sufdir | ruralroute | extra |  city  | state | country | postcode | box |      unit       
----------+-----------+--------+------+---------+------------+---------+--------+------------+-------+--------+-------+---------+----------+-----+-----------------
          | 1         |        |      |         | DEVONSHIRE | PL      |        |            |       | BOSTON | MA    | USA     | 02109    |     | # PENTHOUSE 301
(1 row)

[robe]

It's not supposed to. That's a limitation of extensions you can't force other extensions to install. Only thing I can add is make it required, but I can't do that until I am assured all distributions of postgis will have address_standardizer for postgis 2.2.

[robe]

Just added this link as additional since I had it as #2966 I closed out as dupe

​http://lists.osgeo.org/pipermail/postgis-devel/2014-October/024597.html

[robe]

I assume you mean 9.4beta3 right? These are the counts I get for cali on my windows box. I think they agree with your numbers.

PostgreSQL 9.4beta3, compiled by Visual C++ build 1800, 32-bit POSTGIS="2.2.0de
 r13057" GEOS="3.5.0dev-CAPI-1.9.0 r4026" PROJ="Rel. 4.8.0, 6 March 2012" GDAL=
GDAL 1.11.1, released 2014/09/24" LIBXML="2.7.8" LIBJSON="0.12" RASTER

tabblock I know needs fixing cause I didn't fix it and they changed the struct in tiger 2014.

     table_name     |   cnt
--------------------+---------
 ca_addr            | 2934938
 ca_bg              |   23212
 ca_cousub          |     397
 ca_edges           | 3923238
 ca_faces           | 1133110
 ca_featnames       | 4526071
 ca_place           |    1516
 ca_tabblock        |       0
 ca_tract           |    8057
 ca_zcta5           |    1769
 ca_zip_lookup_base |    4947
 ca_zip_state       |    2214
 ca_zip_state_loc   |    4861

Do you happen to have 2013 data lying around. I don't have for CA but in MA I think the edge count went down presumably because of some fixing. After I loaded MA 2014 data issues that Leo had with geocoding one newly renovated area in cambridge, ma was fixed.

[robe]

My ri numbers match yours too:

     table_name     |  cnt
--------------------+--------
 ri_addr            | 119552
 ri_bg              |    815
 ri_cousub          |     40
 ri_edges           | 130787
 ri_faces           |  38851
 ri_featnames       | 189233
 ri_place           |     34
 ri_tabblock        |      0
 ri_tract           |    244
 ri_zcta5           |     77
 ri_zip_lookup_base |     98
 ri_zip_state       |     95
 ri_zip_state_loc   |     98

[robe]

no issue here except tabblock which is already ticketed in #2915 and I think another