hmap::Tensor Class Reference

A class to represent a multi-dimensional tensor. More...

#include <tensor.hpp>

Collaboration diagram for hmap::Tensor:

Public Member Functions
	Tensor (Vec3< int > shape)
	Construct a new Tensor object.
	Tensor (Vec2< int > shape_xy, int shape_z)
	Construct a new Tensor object.
	Tensor (const std::string &fname)
	Constructs a new Tensor object from a file.
float &	operator() (int i, int j, int k)
	Access an element of the tensor.
const float &	operator() (int i, int j, int k) const
	Access an element of the tensor (const version).
Array	get_slice (int k) const
	Get a 2D slice of the tensor along the z-axis.
float	max () const
	Find the maximum value in the tensor.
float	min () const
	Find the minimum value in the tensor.
void	remap (float vmin=0.f, float vmax=1.f)
	Remap the tensor values to a new range.
Tensor	resample_to_shape_xy (Vec2< int > new_shape_xy)
	Resamples the tensor to a new 2D shape (x, y), shape along z is not changed.
void	set_slice (int k, const Array &slice)
	Set a 2D slice of the tensor along the z-axis.
cv::Mat	to_cv_mat ()
	Convert the tensor to an OpenCV matrix.
std::vector< uint8_t >	to_img_8bit ()
	Convert the tensor to an 8-bit image represented as a vector.
void	to_png (const std::string &fname, int depth=CV_8U)
	Saves the Tensor as a PNG image file.

Public Attributes
Vec3< int >	shape
	Shape of the tensor in 3D space.
std::vector< float >	vector
	Flattened vector containing the tensor's elements.

Detailed Description

A class to represent a multi-dimensional tensor.

The Tensor class is used to store and manipulate multi-dimensional arrays (tensors) of floating-point numbers. It supports element access, basic operations such as finding the minimum and maximum values, remapping the tensor values, and exporting the data to images.

Constructor & Destructor Documentation

Tensor() [1/3]

hmap::Tensor::Tensor

(

Vec3< int >

shape

)

Construct a new Tensor object.

Parameters

shape

3D shape of the tensor.

Tensor() [2/3]

hmap::Tensor::Tensor	(	Vec2< int >	shape_xy,
		int	shape_z
	)

Construct a new Tensor object.

Parameters

shape_xy	2D shape (x, y) of the tensor.
shape_z	Size along the z-axis.

Tensor() [3/3]

hmap::Tensor::Tensor

(

const std::string &

fname

)

Constructs a new Tensor object from a file.

Parameters

fname

The name of the file to load the tensor from.

Member Function Documentation

operator()() [1/2]

float & hmap::Tensor::operator()	(	int	i,
		int	j,
		int	k
	)

Access an element of the tensor.

Parameters

i	Index along the x-axis.
j	Index along the y-axis.
k	Index along the z-axis.

Returns

float& Reference to the element at position (i, j, k).

operator()() [2/2]

const float & hmap::Tensor::operator()	(	int	i,
		int	j,
		int	k
	)		const

Access an element of the tensor (const version).

<This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

Parameters

i	Index along the x-axis.
j	Index along the y-axis.
k	Index along the z-axis.

Returns

const float& Reference to the element at position (i, j, k).

get_slice()

Array hmap::Tensor::get_slice

(

int

k

)

const

Get a 2D slice of the tensor along the z-axis.

Parameters

k	Index along the z-axis.

Returns

An Array representing the 2D slice of the tensor at the specified z-axis index.

max()

float hmap::Tensor::max

(

)

const

Find the maximum value in the tensor.

Returns

float Maximum value in the tensor.

min()

float hmap::Tensor::min

(

)

const

Find the minimum value in the tensor.

Returns

float Minimum value in the tensor.

remap()

void hmap::Tensor::remap	(	float	vmin = 0.f,
		float	vmax = 1.f
	)

Remap the tensor values to a new range.

Parameters

vmin	Minimum value of the new range (default is 0.0).
vmax	Maximum value of the new range (default is 1.0).

resample_to_shape_xy()

Tensor hmap::Tensor::resample_to_shape_xy

(

Vec2< int >

new_shape_xy

)

Resamples the tensor to a new 2D shape (x, y), shape along z is not changed.

Parameters

new_shape_xy

A 2D vector representing the new shape (x, y) of the tensor.

Returns

A new Tensor object with the resampled shape.

set_slice()

void hmap::Tensor::set_slice	(	int	k,
		const Array &	slice
	)

Set a 2D slice of the tensor along the z-axis.

Parameters

k	Index along the z-axis.
slice	The 2D array (slice) to set.

to_cv_mat()

cv::Mat hmap::Tensor::to_cv_mat

(

)

Convert the tensor to an OpenCV matrix.

Returns

cv::Mat OpenCV matrix representing the tensor.

to_img_8bit()

std::vector< uint8_t > hmap::Tensor::to_img_8bit

(

)

Convert the tensor to an 8-bit image represented as a vector.

Returns

std::vector<uint8_t> Vector containing the 8-bit image data.

to_png()

void hmap::Tensor::to_png	(	const std::string &	fname,
		int	depth = CV_8U
	)

Saves the Tensor as a PNG image file.

This function converts the Tensor into a cv::Mat object and then saves it as a PNG image file. The image can be saved in either 8-bit or 16-bit depth, based on the depth parameter. The image is rotated 90 degrees counterclockwise before saving.

Parameters

fname	The filename for the output PNG image, including the path.
depth	The bit depth of the output image. The default is 8-bit (CV_8U). Set to CV_16U for 16-bit output.

Member Data Documentation

shape

Vec3<int> hmap::Tensor::shape

Shape of the tensor in 3D space.

vector

std::vector<float> hmap::Tensor::vector

Flattened vector containing the tensor's elements.

---

The documentation for this class was generated from the following files:

• HighMap/include/highmap/tensor.hpp
• HighMap/src/tensor/tensor.cpp