1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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
|
// Copyright Yahoo. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
#include <vespa/eval/eval/simple_value.h>
#include <vespa/eval/eval/tensor_spec.h>
#include <vespa/eval/eval/value_type.h>
#include <vespa/searchcommon/attribute/config.h>
#include <vespa/searchlib/attribute/attributevector.h>
#include <vespa/searchlib/tensor/distance_calculator.h>
#include <vespa/searchlib/tensor/distance_function_factory.h>
#include <vespa/searchlib/test/attribute_builder.h>
#include <vespa/vespalib/gtest/gtest.h>
#include <iostream>
using namespace search::attribute::test;
using namespace search::attribute;
using namespace search::tensor;
using namespace vespalib::eval;
using search::AttributeVector;
class DistanceCalculatorTest : public testing::Test {
public:
std::shared_ptr<AttributeVector> attr;
const ITensorAttribute* attr_tensor;
std::unique_ptr<DistanceFunction> func;
DistanceCalculatorTest()
: attr(),
attr_tensor(),
func(make_distance_function(DistanceMetric::Euclidean, CellType::DOUBLE))
{
}
void build_attribute(const vespalib::string& tensor_type,
const std::vector<vespalib::string>& tensor_values) {
Config cfg(BasicType::TENSOR);
cfg.setTensorType(ValueType::from_spec(tensor_type));
attr = AttributeBuilder("doc_tensor", cfg).fill_tensor(tensor_values).get();
attr_tensor = dynamic_cast<const ITensorAttribute*>(attr.get());
ASSERT_TRUE(attr_tensor != nullptr);
}
double calc_distance(uint32_t docid, const vespalib::string& query_tensor) {
auto qv = SimpleValue::from_spec(TensorSpec::from_expr(query_tensor));
DistanceCalculator calc(*attr_tensor, *qv, *func);
return calc.calc_with_limit(docid, std::numeric_limits<double>::max());
}
double calc_rawscore(uint32_t docid, const vespalib::string& query_tensor) {
auto qv = SimpleValue::from_spec(TensorSpec::from_expr(query_tensor));
DistanceCalculator calc(*attr_tensor, *qv, *func);
return calc.calc_raw_score(docid);
}
};
constexpr double max_distance = std::numeric_limits<double>::max();
TEST_F(DistanceCalculatorTest, calculation_over_dense_tensor_attribute)
{
build_attribute("tensor(y[2])", {"[3,10]", ""});
vespalib::string qt = "tensor(y[2]):[7,10]";
EXPECT_DOUBLE_EQ(16, calc_distance(1, qt));
EXPECT_DOUBLE_EQ(max_distance, calc_distance(2, qt));
EXPECT_DOUBLE_EQ(1.0/(1.0 + 4.0), calc_rawscore(1, qt));
EXPECT_DOUBLE_EQ(0.0, calc_rawscore(2, qt));
}
TEST_F(DistanceCalculatorTest, calculation_over_mixed_tensor_attribute)
{
build_attribute("tensor(x{},y[2])",
{"{{x:\"a\",y:0}:3,{x:\"a\",y:1}:10,{x:\"b\",y:0}:5,{x:\"b\",y:1}:10}",
"{}", ""});
vespalib::string qt_1 = "tensor(y[2]):[9,10]";
vespalib::string qt_2 = "tensor(y[2]):[1,10]";
EXPECT_DOUBLE_EQ(16, calc_distance(1, qt_1));
EXPECT_DOUBLE_EQ(4, calc_distance(1, qt_2));
EXPECT_DOUBLE_EQ(max_distance, calc_distance(2, qt_1));
EXPECT_DOUBLE_EQ(max_distance, calc_distance(3, qt_1));
EXPECT_DOUBLE_EQ(1.0/(1.0 + 4.0), calc_rawscore(1, qt_1));
EXPECT_DOUBLE_EQ(1.0/(1.0 + 2.0), calc_rawscore(1, qt_2));
EXPECT_DOUBLE_EQ(0.0, calc_rawscore(2, qt_1));
EXPECT_DOUBLE_EQ(0.0, calc_rawscore(3, qt_1));
}
GTEST_MAIN_RUN_ALL_TESTS()
|
