neighbor 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f346d121745bd39998267c45f77009970c014de887376022f8b6425192d37354
-  data.tar.gz: '0293bc9fd3633ca3ee811940c59995407357d4b880fcf8ea8cd9223e5df0e75b'
+  metadata.gz: 01df817b036ee9b4d0c54ddadfb3be7fb31b36bb81504321b8dfd3d7a1ba83a4
+  data.tar.gz: f7a1757914f2e1226bc3b95fe5e0df9837a530315d6e627062df565e331b1708
 SHA512:
-  metadata.gz: cf080c46ad8133a460453c773faabac02e229821625fa9f2d51a320195b1e821b230c94c7fe45bc7f7d12b0b51ce6d3e5f1d4c0d74e446b0ce22dde2f5171cde
-  data.tar.gz: f2e0ae2247979bd3dd083b1869b56f7b6ed879ce8dd1c0ef31d965392d8250548acb304a78ba99a1d6cdfea116c10e0d67aff9c06185a793e576bddae4d624f1
+  metadata.gz: 8381834f92092cb13d2c8898b588f0758e99e48f84e3b506aab15acf54fb0b3f72449734ca68db10de466eecde4a7b277cceca72192240ce21567125ad0af52f
+  data.tar.gz: 79ddd953bf0c134d73c92c61a8051524c024e2e1c95710e346a191a8a7c098f55562f81263752ea2e2ccaef9ced8e376dbf6431a4e745931d256ee2ca854d309

data/CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.1.1 (2021-02-16)
+
+- Fixed `Could not dump table` error
+
 ## 0.1.0 (2021-02-15)
 
 - First release

data/README.md

@@ -20,6 +20,8 @@ rails generate neighbor:install
 rails db:migrate
 ```
 
+This enables the [cube extension](https://www.postgresql.org/docs/current/cube.html) in Postgres
+
 ## Getting Started
 
 Create a migration
@@ -48,13 +50,13 @@ item.update(neighbor_vector: [1.0, 1.2, 0.5])
 
 > With cosine distance (the default), vectors are normalized before being stored
 
-And get the nearest neighbors
+Get the nearest neighbors
 
 ```ruby
 item.nearest_neighbors.first(5)
 ```
 
-## Distances
+## Distance
 
 Specify the distance metric
 
@@ -64,19 +66,24 @@ class Item < ApplicationRecord
 end
 ```
 
-Supported distances are:
+Supported values are:
 
 - `cosine` (default)
 - `euclidean`
 - `taxicab`
 - `chebyshev`
 
-Returned records will have a `neighbor_distance` attribute
+Records returned from `nearest_neighbors` will have a `neighbor_distance` attribute
 
 ```ruby
-returned_item.neighbor_distance
+nearest_item = item.nearest_neighbors.first
+nearest_item.neighbor_distance
 ```
 
+## Dimensions
+
+By default, Postgres limits the `cube` data type to 100 dimensions. See the [Postgres docs](https://www.postgresql.org/docs/current/cube.html) for how to increase this.
+
 ## Example
 
 You can use Neighbor for online item recommendations with [Disco](https://github.com/ankane/disco). We’ll use MovieLens data for this example.
@@ -119,6 +126,8 @@ movie = Movie.find_by(name: "Star Wars (1977)")
 movie.nearest_neighbors.first(5).map(&:name)
 ```
 
+[Complete code](examples/disco.rb)
+
 ## History
 
 View the [changelog](https://github.com/ankane/neighbor/blob/master/CHANGELOG.md)

data/lib/neighbor.rb

@@ -21,7 +21,20 @@ ActiveSupport.on_load(:active_record) do
 
   extend Neighbor::Model
 
-  # prevent unknown OID warning
   require "active_record/connection_adapters/postgresql_adapter"
+
+  # ensure schema can be dumped
+  ActiveRecord::ConnectionAdapters::PostgreSQLAdapter::NATIVE_DATABASE_TYPES[:cube] = {name: "cube"}
+
+  # ensure schema can be loaded
+  if ActiveRecord::VERSION::MAJOR >= 6
+    ActiveRecord::ConnectionAdapters::TableDefinition.send(:define_column_methods, :cube)
+  else
+    ActiveRecord::ConnectionAdapters::TableDefinition.define_method :cube do |*args, **options|
+      args.each { |name| column(name, :cube, options) }
+    end
+  end
+
+  # prevent unknown OID warning
   ActiveRecord::ConnectionAdapters::PostgreSQLAdapter.prepend(Neighbor::RegisterCubeType)
 end

data/lib/neighbor/model.rb

@@ -4,8 +4,12 @@ module Neighbor
       distance = distance.to_s
       raise ArgumentError, "Invalid distance: #{distance}" unless %w(cosine euclidean taxicab chebyshev).include?(distance)
 
+      # TODO make configurable
+      # likely use argument
+      attribute_name = :neighbor_vector
+
       class_eval do
-        attribute :neighbor_vector, Neighbor::Vector.new(dimensions: dimensions, distance: distance)
+        attribute attribute_name, Neighbor::Vector.new(dimensions: dimensions, distance: distance)
 
         define_method :nearest_neighbors do
           return self.class.none if neighbor_vector.nil?
@@ -20,9 +24,11 @@ module Neighbor
               "<->"
             end
 
+          quoted_attribute = "#{self.class.connection.quote_table_name(self.class.table_name)}.#{self.class.connection.quote_column_name(attribute_name)}"
+
           # important! neighbor_vector should already be typecast
           # but use to_f as extra safeguard against SQL injection
-          order = "neighbor_vector #{operator} cube(array[#{neighbor_vector.map(&:to_f).join(", ")}])"
+          order = "#{quoted_attribute} #{operator} cube(array[#{neighbor_vector.map(&:to_f).join(", ")}])"
 
           # https://stats.stackexchange.com/questions/146221/is-cosine-similarity-identical-to-l2-normalized-euclidean-distance
           # with normalized vectors:
@@ -35,7 +41,7 @@ module Neighbor
           self.class
             .select(*self.class.column_names, "#{neighbor_distance} AS neighbor_distance")
             .where.not(self.class.primary_key => send(self.class.primary_key))
-            .where.not(neighbor_vector: nil)
+            .where.not(attribute_name => nil)
             .order(Arel.sql(order))
         end
       end

data/lib/neighbor/vector.rb

@@ -13,11 +13,7 @@ module Neighbor
       raise Error, "Expected #{@dimensions} dimensions, not #{value.size}" unless value.size == @dimensions
 
       if @distance == "cosine"
-        norm = 0.0
-        value.each do |v|
-          norm += v * v
-        end
-        norm = Math.sqrt(norm)
+        norm = Math.sqrt(value.sum { |v| v * v })
         value.map { |v| v / norm }
       else
         value

data/lib/neighbor/version.rb

@@ -1,3 +1,3 @@
 module Neighbor
-  VERSION = "0.1.0"
+  VERSION = "0.1.1"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: neighbor
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Andrew Kane
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-02-16 00:00:00.000000000 Z
+date: 2021-02-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activerecord
@@ -58,7 +58,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.4
+rubygems_version: 3.2.3
 signing_key:
 specification_version: 4
 summary: Nearest neighbor search for Rails and Postgres