Module: Neighbor::SQLite

Defined in:
lib/neighbor/sqlite.rb

Defined Under Namespace

Modules: InstanceMethods

Class Attribute Summary collapse

Class Method Summary collapse

Class Attribute Details

.extensions ⇒ Object (readonly)

Returns the value of attribute extensions.



4
5
6
 
# File 'lib/neighbor/sqlite.rb', line 4

def extensions
  @extensions
end

Class Method Details

.initialize!(extension: :sqlite_vec) ⇒ Object

note: this is a public API (unlike PostgreSQL and MySQL)



8
9
10
11
12
13
14
15
16
 
# File 'lib/neighbor/sqlite.rb', line 8

def self.initialize!(extension: :sqlite_vec)
  if extension == :sqlite_vec
    require "sqlite_vec"
  elsif !extension.is_a?(String)
    raise ArgumentError, "Unsupported extension"
  end

  (@extensions ||= []) << extension
end

.initialize_adapter! ⇒ Object 



18
19
20
21
22
23
24
25
26
 
# File 'lib/neighbor/sqlite.rb', line 18

def self.initialize_adapter!
  @extensions ||= []

  require_relative "type/sqlite_vector"
  require_relative "type/sqlite_int8_vector"

  require "active_record/connection_adapters/sqlite3_adapter"
  ActiveRecord::ConnectionAdapters::SQLite3Adapter.prepend(InstanceMethods)
end

.setup_functions(db) ⇒ Object 



36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
 
# File 'lib/neighbor/sqlite.rb', line 36

def self.setup_functions(db)
  db.create_function("neighbor_l2_distance", 2) do |func, a, b, c|
    func.result =
      if a.nil? || b.nil?
        nil
      else
        raise SQLite3::SQLException, "different vector dimensions" if a.bytesize != b.bytesize
        fmt = c == 1 ? "c*" : "f*"
        a = a.unpack(fmt)
        b = b.unpack(fmt)
        Math.sqrt(a.zip(b).sum { |ai, bi| (ai - bi)**2 })
      end
  end

  db.create_function("neighbor_max_inner_product", 2) do |func, a, b, c|
    func.result =
      if a.nil? || b.nil?
        nil
      else
        raise SQLite3::SQLException, "different vector dimensions" if a.bytesize != b.bytesize
        fmt = c == 1 ? "c*" : "f*"
        a = a.unpack(fmt)
        b = b.unpack(fmt)
        -a.zip(b).sum { |ai, bi| ai * bi }
      end
  end

  db.create_function("neighbor_cosine_distance", 2) do |func, a, b, c|
    func.result =
      if a.nil? || b.nil?
        nil
      else
        raise SQLite3::SQLException, "different vector dimensions" if a.bytesize != b.bytesize
        fmt = c == 1 ? "c*" : "f*"
        a = a.unpack(fmt)
        b = b.unpack(fmt)
        similarity = a.zip(b).sum { |ai, bi| ai * bi }
        norma = a.sum { |v| v * v }
        normb = b.sum { |v| v * v }
        1.0 - (similarity / Math.sqrt(norma * normb)).clamp(-1.0, 1.0)
      end
  end

  db.create_function("neighbor_l1_distance", 2) do |func, a, b, c|
    func.result =
      if a.nil? || b.nil?
        nil
      else
        raise SQLite3::SQLException, "different vector dimensions" if a.bytesize != b.bytesize
        fmt = c == 1 ? "c*" : "f*"
        a = a.unpack(fmt)
        b = b.unpack(fmt)
        a.zip(b).sum { |ai, bi| (ai - bi).abs }
      end
  end

  db.create_function("neighbor_hamming_distance", 2) do |func, a, b|
    func.result =
      if a.nil? || b.nil?
        nil
      else
        raise SQLite3::SQLException, "different vector dimensions" if a.bytesize != b.bytesize
        # TODO improve
        a.each_byte.zip(b.each_byte).sum { |ai, bi| (ai ^ bi).to_s(2).count("1") }
      end
  end

  db.create_function("neighbor_jaccard_distance", 2) do |func, a, b|
    func.result =
      if a.nil? || b.nil?
        nil
      else
        raise SQLite3::SQLException, "different vector dimensions" if a.bytesize != b.bytesize
        # TODO improve
        ab = a.each_byte.zip(b.each_byte).sum { |ai, bi| (ai & bi).to_s(2).count("1") }
        aa = a.unpack1("B*").count("1")
        bb = b.unpack1("B*").count("1")
        ab == 0 ? 1.0 : 1.0 - (ab / (aa + bb - ab).to_f)
      end
  end
end

.sqlite_vec? ⇒ Boolean

Returns:

  • (Boolean) 


32
33
34
 
# File 'lib/neighbor/sqlite.rb', line 32

def self.sqlite_vec?
  extensions.include?(:sqlite_vec)
end

.vec1? ⇒ Boolean

Returns:

  • (Boolean) 


28
29
30
 
# File 'lib/neighbor/sqlite.rb', line 28

def self.vec1?
  extensions.any?(String)
end