Add comments

README.md

@@ -7,6 +7,8 @@ CartoDB Spatial Analysis extension for PostgreSQL.
 * *pg* contains the PostgreSQL extension source code
 * *python* Python module
 
+FIXME: should it be `./extension` and `./lib/python' ?
+
 ## Requirements
 
 * pip

pg/crankshaft--0.0.1.sql

@@ -1,6 +1,3 @@
---DO NOT MODIFY THIS FILE, IT IS GENERATED AUTOMATICALLY FROM SOURCES
--- Complain if script is sourced in psql, rather than via CREATE EXTENSION
-\echo Use "CREATE EXTENSION crankshaft" to load this file. \quit
 -- Internal function.
 -- Set the seeds of the RNGs (Random Number Generators)
 -- used internally.
@@ -136,13 +133,4 @@ BEGIN
   RETURN ST_Collect(points);
 END;
 $$
-LANGUAGE plpgsql VOLATILE;
--- Make sure by default there are no permissions for publicuser
--- NOTE: this happens at extension creation time, as part of an implicit transaction.
--- REVOKE ALL PRIVILEGES ON SCHEMA cdb_crankshaft FROM PUBLIC, publicuser CASCADE;
-
--- Grant permissions on the schema to publicuser (but just the schema)
-GRANT USAGE ON SCHEMA cdb_crankshaft TO publicuser;
-
--- Revoke execute permissions on all functions in the schema by default
--- REVOKE EXECUTE ON ALL FUNCTIONS IN SCHEMA cdb_crankshaft FROM PUBLIC, publicuser;
+LANGUAGE plpgsql VOLATILE

pg/sql/0.0.1/00_header.sql

@@ -1,3 +0,0 @@
---DO NOT MODIFY THIS FILE, IT IS GENERATED AUTOMATICALLY FROM SOURCES
--- Complain if script is sourced in psql, rather than via CREATE EXTENSION
-\echo Use "CREATE EXTENSION crankshaft" to load this file. \quit

pg/sql/0.0.1/04_dot_density.sql

@@ -51,4 +51,4 @@ BEGIN
   RETURN ST_Collect(points);
 END;
 $$
-LANGUAGE plpgsql VOLATILE;
+LANGUAGE plpgsql VOLATILE

pg/sql/0.0.1/90_permissions.sql

@@ -1,9 +0,0 @@
--- Make sure by default there are no permissions for publicuser
--- NOTE: this happens at extension creation time, as part of an implicit transaction.
--- REVOKE ALL PRIVILEGES ON SCHEMA cdb_crankshaft FROM PUBLIC, publicuser CASCADE;
-
--- Grant permissions on the schema to publicuser (but just the schema)
-GRANT USAGE ON SCHEMA cdb_crankshaft TO publicuser;
-
--- Revoke execute permissions on all functions in the schema by default
--- REVOKE EXECUTE ON ALL FUNCTIONS IN SCHEMA cdb_crankshaft FROM PUBLIC, publicuser;

pg/sql/0.0.1/population.sql

@@ -0,0 +1,138 @@
+-- Function to obtain an estimate of the population living inside
+-- an area (polygon) from the CartoDB Data Observatory
+CREATE OR REPLACE FUNCTION cdb_population(area geometry)
+RETURNS NUMERIC AS $$
+DECLARE
+  georef_column TEXT;
+  table_id TEXT;
+  tag_value TEXT;
+  table_name TEXT;
+  column_name TEXT;
+  population NUMERIC;
+BEGIN
+
+  -- Note: comments contain pseudo-code that should be implemented
+
+  -- Register metadata tables:
+  -- This would require super-user privileges
+  /*
+  SELECT cdb_add_remote_table('observatory', 'bmd_column_table');
+  SELECT cdb_add_remote_table('observatory', 'bmd_column_2_column');
+  SELECT cdb_add_remote_table('observatory', 'bmd_table');
+  SELECT cdb_add_remote_table('observatory', 'bmd_column_table');
+  SELECT cdb_add_remote_table('observatory', 'bmd_column_tag');
+  SELECT cdb_add_remote_table('observatory', 'bmd_tag');
+  */
+
+  tag_value := 'population';
+
+
+  -- Determine the georef column id to be used: it must have type 'geometry',
+  -- the maximum weight.
+  -- TODO: in general, multiple columns with maximal weight could be found;
+  -- we should use the timespan of the table to disambiguate (choose the
+  -- most recent). Also a rank of geometry columns should be introduced to
+  -- find select the greatest resolution available.
+  /*
+  WITH selected_tables AS (
+    -- Find tables that have population columns and cover the input area
+    SELECT tab.id AS id
+    FROM observatory.bmd_column col,
+         observatory.bmd_column_table coltab,
+         observatory.bmd_table tab,
+         observatory.bmd_tag tag,
+         observatory.bmd_column_tag coltag
+    WHERE coltab.column_id = col.id
+      AND coltab.table_id = tab.id
+      AND coltag.tag_id = tag.id
+      AND coltag.column_id = col.id
+      AND tag.name ILIKE tag_value
+      AND tab.id = table_id
+      AND tab.bounds && area;
+  )
+  SELECT
+    FROM bmd_column col
+    JOIN bmd_table tab ON col.table_id = tab.id
+    WHERE type = 'geometry'
+      AND tab.id IN (selected_tables)
+    ORDER BY weight DESC LIMIT 1;
+  */
+  georef_column := '"us.census.tiger".block_group_2013';
+
+  -- Now we will query the metadata to find which actual tables correspond
+  -- to this datasource and resolution/timespan
+  -- and choose the 'parent' or more general of them.
+  /*
+  SELECT from_table_geoid.id data_table_id
+  FROM observatory.bmd_column_table from_column_table_geoid,
+       observatory.bmd_column_table to_column_table_geoid,
+       observatory.bmd_column_2_column rel,
+       observatory.bmd_column_table to_column_table_geom,
+       observatory.bmd_table from_table_geoid,
+       observatory.bmd_table to_table_geoid,
+       observatory.bmd_table to_table_geom
+  WHERE from_column_table_geoid.column_id = to_column_table_geoid.column_id
+    AND to_column_table_geoid.column_id = rel.from_id
+    AND rel.reltype = 'geom_ref'
+    AND rel.to_id = to_column_table_geom.column_id
+    AND to_column_table_geom.column_id = georef_column
+    AND from_table_geoid.id = from_column_table_geoid.table_id
+    AND to_table_geoid.id = to_column_table_geoid.table_id
+    AND to_table_geom.id = to_column_table_geom.table_id
+    AND from_table_geoid.bounds && area
+  ORDER by from_table_geoid.timespan desc
+  INTO table_id;
+  */
+  table_id := '"us.census.acs".extract_2013_5yr_block_group';
+
+  -- Next will fetch the columns of that table that are tagged as population:
+  -- and get the more general one (not having a parent or denominator)
+  /*
+  WITH column_ids AS (
+    SELECT col.id AS id
+    FROM observatory.bmd_column col,
+         observatory.bmd_column_table coltab,
+         observatory.bmd_table tab,
+         observatory.bmd_tag tag,
+         observatory.bmd_column_tag coltag
+    WHERE coltab.column_id = col.id
+      AND coltab.table_id = tab.id
+      AND coltag.tag_id = tag.id
+      AND coltag.column_id = col.id
+      AND tag.name ILIKE tag_value
+      AND tab.id = table_id;
+  ),
+  excluded_column_ids AS (
+    SELECT from_id AS id
+    FROM observatory.bmd_column_2_column
+    WHERE from_id in (column_ids)
+      AND reltype in ('parent', 'denominator')
+      AND to_id in (column_ids)
+  ),
+  SELECT bmd_table.tablename, bmd_column_table.colname
+  FROM observatory.bmd_column_table,
+       observatory.bmd_table
+  WHERE bmd_column_table.table_id = bmd_table.id
+    AND bmd_column_table.column_id IN (column_ids)
+    AND NOT bmd_column_table.column_id IN (exclude_column_ids)
+  INTO (table_name, column_name);
+  */
+  table_name := 'us_census_acs2013_5yr_block_group';
+  column_name := 'total_pop';
+
+  -- Register the foreign table
+  -- This would require super-user privileges
+  -- SELECT cdb_add_remote_table('observatory', table_name);
+
+  -- Perform the query
+  SELECT cdb_crankshaft.cdb_overlap_sum(
+    area,
+    table_name,
+    column_name,
+    schema_name := 'observatory')
+  INTO population;
+
+  RETURN population;
+END;
+$$
+LANGUAGE plpgsql VOLATILE

pg/test/0.0.1/sql/90_permissions.sql

@@ -1,18 +0,0 @@
-SELECT cdb_crankshaft._cdb_random_seeds(1234);
-
--- Use regular user role
-SET ROLE test_regular_user;
-
--- Add to the search path the schema
-SET search_path TO public,cartodb,cdb_crankshaft;
-
--- Exercise public functions
-SELECT ppoints.code, m.quads
-  FROM ppoints
-  JOIN cdb_moran_local('ppoints', 'value') m
-    ON ppoints.cartodb_id = m.ids
-  ORDER BY ppoints.code;
-SELECT round(cdb_overlap_sum(
-  '0106000020E61000000100000001030000000100000004000000FFFFFFFFFF3604C09A0B9ECEC42E444000000000C060FBBF30C7FD70E01D44400000000040AD02C06481F1C8CD034440FFFFFFFFFF3604C09A0B9ECEC42E4440'::geometry,
-  'values', 'value'
-), 2);

python/crankshaft/setup.py

@@ -40,9 +40,9 @@ setup(
 
     # The choice of component versions is dictated by what's
     # provisioned in the production servers.
-    install_requires=['pysal==1.11.0', 'numpy==1.10.4', 'scipy==0.17.0', 'pandas==0.17.1', 'scikit-learn==0.17.0', 'statsmodels==0.6.1', 'keras==0.3.2', 'shapely==1.5.3', 'osgeo==2.0.2','scikit-image==0.12.3'],
+    install_requires=['pysal==1.11.0','numpy==1.6.1','scipy==0.17.0'],
 
-    requires=['pysal', 'numpy', 'scipy', 'pandas', 'scikit-learn', 'statsmodels', 'keras', 'shapely', 'osgeo','skimage'],
+    requires=['pysal', 'numpy'],
 
     test_suite='test'
 )