sqlite/ext/vec1/test/vec1ac.test

# 2026-01-17
#
# The author disclaims copyright to this source code.  In place of
# a legal notice, here is a blessing:
#
#    May you do good and not evil.
#    May you find forgiveness for yourself and forgive others.
#    May you share freely, never taking more than you give.
#
#***********************************************************************
#
#

source [file join [file dirname [info script]] vec1_common.tcl]
set testprefix vec1ac
vec1_load db

db func normalize normalize
proc normalize {vec} {
  binary scan $vec f* aVal
  set sum 0.0
  foreach v $aVal {
    set sum [expr $sum + ($v*$v)]
  }
  set sz [expr sqrt($sum)]
  set aNew [list]
  foreach v $aVal {
    lappend aNew [expr $v/$sz]
  }
  binary format f* $aNew
}

db func rescale rescale
proc rescale {vec mul} {
  binary scan $vec f* aVal
  set aNew [list]
  foreach v $aVal {
    lappend aNew [expr $v*$mul]
  }
  binary format f* $aNew
}

do_execsql_test 1.0 {
  CREATE TABLE t1(k INTEGER PRIMARY KEY, v VECTOR);
  WITH s(i) AS (
    SELECT 1 UNION ALL SELECT i+1 FROM s WHERE i<10000
  )
  INSERT INTO t1 SELECT NULL, random_vector2(128) FROM s;

  CREATE VIRTUAL TABLE v1 USING vec1;
  INSERT INTO v1(rowid, vector) SELECT k, v FROM t1;
}

foreach {tn cfg} {
  1 { {nbucket:0, codesize:0, distance:"cos"} }
  2 { {nbucket:0, codesize:32, distance:"cos"} }
  3 { {nbucket:32, codesize:32, distance:"cos"} }
} {

  do_execsql_test 1.$tn.1 {
    INSERT INTO v1(cmd, vector) VALUES('rebuild', (
      SELECT vec1_train(vector, $cfg) FROM v1
    ));
  } {}

  do_execsql_test 1.$tn.2 {
    SELECT sum( 
      rowid IN ( SELECT rowid FROM v1( normalize( o.vector ), '{K:10}' ) ) 
    )
    FROM v1 AS o WHERE rowid BETWEEN 600 AND 700;
  } {101}

  do_execsql_test 1.$tn.3 {
    SELECT sum( 
      rowid IN ( SELECT rowid FROM v1( rescale( o.vector, 0.01 ), '{K:10}' ) ) 
    )
    FROM v1 AS o WHERE rowid BETWEEN 600 AND 700;
  } {101}
}

do_execsql_test 2.0 {
  CREATE VIRTUAL TABLE v2 USING vec1;
  INSERT INTO v2(rowid, vector) SELECT k, v FROM t1;
  INSERT INTO v2(cmd, vector) VALUES('rebuild', (
      SELECT vec1_train(vector, '{nbucket:32, codesize:32, distance:"cos"}') 
      FROM v2
  ));
}

do_execsql_test 2.1 {
  PRAGMA integrity_check;
} {ok}

do_execsql_test 2.2 {
  DELETE FROM v2 WHERE (rowid % 45)==0;
  PRAGMA integrity_check;
} {ok}


finish_test