Collaboration diagram for Float Distances:

Modules
	Dynamic Time Warping Distance

Macros
#define	Ln2 __logf_data.ln2
	Jensen-Shannon distance between two vectors. More...

#define	BITS_N (1 << LOGF_TABLE_BITS)

#define	OFF 0x3f330000

Functions
f32_t	tpt_canberra_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, uint32_t aCount)
	Canberra distance between two vectors. More...

f32_t	tpt_chebyshev_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, uint32_t aCount)
	Chebyshev distance between two vectors. More...

f32_t	tpt_cityblock_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, uint32_t aCount)
	Cityblock (Manhattan) distance between two vectors. More...

f32_t	tpt_correlation_distance_f32 (f32_t aInDataA, f32_t aInDataB, uint32_t aCount)
	Correlation distance between two vectors. More...

f32_t	tpt_cosine_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, uint32_t aCount)
	Cosine distance between two vectors. More...

f64_t	tpt_cosine_distance_f64 (const f64_t aInDataA, const f64_t aInDataB, uint32_t aCount)
	Cosine distance between two vectors. More...

f32_t	tpt_euclidean_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, uint32_t aCount)
	Euclidean distance between two vectors. More...

f64_t	tpt_euclidean_distance_f64 (const f64_t aInDataA, const f64_t aInDataB, uint32_t aCount)
	Euclidean distance between two vectors. More...

static f32_t	_logf_ (f32_t x)

f32_t	tpt_jensenshannon_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, uint32_t aCount)
	Jensen-Shannon distance between two vectors. More...

f32_t	tpt_minkowski_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, int32_t aOrder, uint32_t aCount)
	Minkowski distance between two vectors. More...

Variables
const struct logf_data	__logf_data

f32_t	tpt_braycurtis_distance_f32 (const f32_t aInDataA, const f32_t aInDataB, uint32_t aCount)
	Bray-Curtis distance between two vectors. More...

Detailed Description

Distances between two vectors of float values.

Macro Definition Documentation

◆ BITS_N

#define BITS_N (1 << LOGF_TABLE_BITS)

◆ Ln2

#define Ln2 __logf_data.ln2

Jensen-Shannon distance between two vectors.

This function is assuming that elements of second vector are > 0 and 0 only when the corresponding element of first vector is 0. Otherwise the result of the computation does not make sense and for speed reasons, the cases returning NaN or Infinity are not managed.

When the function is computing x log (x / y) with x == 0 and y == 0, it will compute the right result (0) but a division by zero will occur and should be ignored in client code.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ OFF

#define OFF 0x3f330000

Function Documentation

◆ _logf_()

static f32_t _logf_ ( f32_t x )

inlinestatic

◆ tpt_braycurtis_distance_f32()

f32_t tpt_braycurtis_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		uint32_t	aCount
	)

Bray-Curtis distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_canberra_distance_f32()

f32_t tpt_canberra_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		uint32_t	aCount
	)

Canberra distance between two vectors.

This function may divide by zero when samples aInDataA[i] and aInDataB[i] are both zero. The result of the computation will be correct. So the division per zero may be ignored.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_chebyshev_distance_f32()

f32_t tpt_chebyshev_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		uint32_t	aCount
	)

Chebyshev distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_cityblock_distance_f32()

f32_t tpt_cityblock_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		uint32_t	aCount
	)

Cityblock (Manhattan) distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_correlation_distance_f32()

f32_t tpt_correlation_distance_f32	(	f32_t *	aInDataA,
		f32_t *	aInDataB,
		uint32_t	aCount
	)

Correlation distance between two vectors.

The input vectors are modified in place!

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_cosine_distance_f32()

f32_t tpt_cosine_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		uint32_t	aCount
	)

Cosine distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_cosine_distance_f64()

f64_t tpt_cosine_distance_f64	(	const f64_t *	aInDataA,
		const f64_t *	aInDataB,
		uint32_t	aCount
	)

Cosine distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_euclidean_distance_f32()

f32_t tpt_euclidean_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		uint32_t	aCount
	)

Euclidean distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_euclidean_distance_f64()

f64_t tpt_euclidean_distance_f64	(	const f64_t *	aInDataA,
		const f64_t *	aInDataB,
		uint32_t	aCount
	)

Euclidean distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_jensenshannon_distance_f32()

f32_t tpt_jensenshannon_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		uint32_t	aCount
	)

Jensen-Shannon distance between two vectors.

This function is assuming that elements of second vector are > 0 and 0 only when the corresponding element of first vector is 0. Otherwise the result of the computation does not make sense and for speed reasons, the cases returning NaN or Infinity are not managed.

When the function is computing x log (x / y) with x == 0 and y == 0, it will compute the right result (0) but a division by zero will occur and should be ignored in client code.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aCount	vector length

Returns: distance

◆ tpt_minkowski_distance_f32()

f32_t tpt_minkowski_distance_f32	(	const f32_t *	aInDataA,
		const f32_t *	aInDataB,
		int32_t	aOrder,
		uint32_t	aCount
	)

Minkowski distance between two vectors.

Parameters

[in]	aInDataA	First vector
[in]	aInDataB	Second vector
[in]	aOrder	Distance order
[in]	aCount	Number of samples

Returns: distance

Variable Documentation

◆ __logf_data

const struct logf_data __logf_data

Initial value:

= {
    .tab =
        {
            {0x1.661ec79f8f3bep+0, -0x1.57bf7808caadep-2},
            {0x1.571ed4aaf883dp+0, -0x1.2bef0a7c06ddbp-2},
            {0x1.49539f0f010bp+0, -0x1.01eae7f513a67p-2},
            {0x1.3c995b0b80385p+0, -0x1.b31d8a68224e9p-3},
            {0x1.30d190c8864a5p+0, -0x1.6574f0ac07758p-3},
            {0x1.25e227b0b8eap+0, -0x1.1aa2bc79c81p-3},
            {0x1.1bb4a4a1a343fp+0, -0x1.a4e76ce8c0e5ep-4},
            {0x1.12358f08ae5bap+0, -0x1.1973c5a611cccp-4},
            {0x1.0953f419900a7p+0, -0x1.252f438e10c1ep-5},
            {0x1p+0, 0x0p+0},
            {0x1.e608cfd9a47acp-1, 0x1.aa5aa5df25984p-5},
            {0x1.ca4b31f026aap-1, 0x1.c5e53aa362eb4p-4},
            {0x1.b2036576afce6p-1, 0x1.526e57720db08p-3},
            {0x1.9c2d163a1aa2dp-1, 0x1.bc2860d22477p-3},
            {0x1.886e6037841edp-1, 0x1.1058bc8a07ee1p-2},
            {0x1.767dcf5534862p-1, 0x1.4043057b6ee09p-2},
        },
    .ln2 = 0x1.62e42fefa39efp-1,
    .poly = {
        -0x1.00ea348b88334p-2,
        0x1.5575b0be00b6ap-2,
        -0x1.ffffef20a4123p-2,
    }}

Modules

Macros

Functions

Variables

Detailed Description

Macro Definition Documentation

◆ BITS_N

◆ Ln2

◆ OFF

Function Documentation

◆ _logf_()

◆ tpt_braycurtis_distance_f32()

◆ tpt_canberra_distance_f32()

◆ tpt_chebyshev_distance_f32()

◆ tpt_cityblock_distance_f32()

◆ tpt_correlation_distance_f32()

◆ tpt_cosine_distance_f32()

◆ tpt_cosine_distance_f64()

◆ tpt_euclidean_distance_f32()

◆ tpt_euclidean_distance_f64()

◆ tpt_jensenshannon_distance_f32()

◆ tpt_minkowski_distance_f32()

Variable Documentation

◆ __logf_data