porescene.utility module

class porescene.utility.CompassDirection(*values)

Bases: Enum

EAST = 'E'

NORTH = 'N'

NORTHEAST = 'NE'

NORTHWEST = 'NW'

SOUTH = 'S'

SOUTHEAST = 'SE'

SOUTHWEST = 'SW'

WEST = 'W'

class porescene.utility.Mesh(vertices, faces, name='object')

Bases: object

A simple container pairing a mesh’s vertices with its faces.

property faces: ndarray

(M, K) array of vertex indices, one row per polygon.

property name: str

Name of the mesh.

property vertices: ndarray

(N, 3) array of vertex coordinates.

class porescene.utility.MultiplicationSymbol(*values)

Bases: Enum

CROSS = '×'

DOT = '·'

class porescene.utility.Orientation(*values)

Bases: Enum

HORIZONTAL = 'H'

VERTICAL = 'V'

class porescene.utility.UnitExponentMetric(*values)

Bases: Enum

ATTO = -18

CENTI = -2

DECA = 1

DECI = -1

EXA = 18

FEMTO = -15

GIGA = 9

HECTO = 2

KILO = 3

MEGA = 6

MICRO = -6

MILLI = -3

NANO = -9

PETA = 15

PICO = -12

QUECTO = -30

QUETTA = 30

RONNA = 27

RONTO = -27

TERA = 12

YOCTO = -24

YOTTA = 24

ZEPTO = -21

ZETTA = 21

class porescene.utility.UnitPrefixMetric(*values)

Bases: Enum

ATTO = 'a'

CENTI = 'c'

DECA = 'da'

DECI = 'd'

EXA = 'E'

FEMTO = 'f'

GIGA = 'G'

HECTO = 'h'

KILO = 'k'

MEGA = 'M'

MICRO = 'µ'

MILLI = 'm'

NANO = 'n'

PETA = 'P'

PICO = 'p'

QUECTO = 'q'

QUETTA = 'Q'

RONNA = 'R'

RONTO = 'r'

TERA = 't'

YOCTO = 'y'

YOTTA = 'Y'

ZEPTO = 'z'

ZETTA = 'Z'

porescene.utility.n_equidistant(lst, n)

porescene.utility.suppress_stdout()

Silences stdout at the OS file-descriptor level.

Return type:

Generator[None]

porescene.utility.svg2png(pth, crop=True)

Convert a SVG file to PNG.

Uses resvg_py for the conversion, which ships as a self-contained wheel, so pip install pulls in everything and no external software is required. With crop=True the result is trimmed to its visible content by removing the surrounding transparent margin.

Note that resvg_py only renders content inside the SVG viewport; anything drawn beyond the root <svg> width/height is clipped before the crop and cannot be recovered.

Parameters:

• pth (Path) – Path to the file to be converted.

• crop (bool) – Trim the PNG to the bounding box of its non-transparent pixels.

Return type:

Path

porescene.utility.volume2mesh(img, voxel_size=(1.0, 1.0, 1.0), labels=1, *, per_label=False, swap_axes=False, name='object')

Overloads:

• img (np.ndarray), voxel_size (float | Sequence[float]), labels (int | Sequence[int] | np.ndarray), per_label (Literal[False]), swap_axes (bool), name (str) → Mesh

• img (np.ndarray), voxel_size (float | Sequence[float]), labels (int | Sequence[int] | np.ndarray), per_label (Literal[True]), swap_axes (bool), name (str) → dict[int, Mesh]

Builds a rectangular (quad) surface mesh from a labeled voxel image.

Extracts the axis-aligned boundary faces of the voxels whose label is included in labels – every face between a selected voxel and a non-selected neighbour (or the volume border) – as unit squares scaled by the voxel size. Coincident vertices are merged, so each position is stored once and shared between faces.

Parameters:

• img (ndarray) – 2D or 3D integer array of voxel labels. A 2D image is treated as a single slice, one voxel thick along the third axis.

• voxel_size (float | Sequence[float]) – Edge length of a voxel. A scalar applies to all three axes; a 3-element sequence gives the length along x, y and z, by default (1.0, 1.0, 1.0).

• labels (int | Sequence[int] | ndarray) – Label value(s) to include in the mesh, by default 1.

• per_label (bool) – If False (default), all selected labels are meshed into a single surface. If True, a separate mesh is built per label and a {label: Mesh} mapping is returned.

• swap_axes (bool) – If True, swaps the first and third axes of img before meshing, leaving the second axis untouched, by default False. This compensates for the axis-order mismatch between MATLAB’s column-major and numpy’s row-major array storage (see the corresponding swap_axes option of porescene.model.PoreNetwork.from_mat()): a voxel image loaded straight from a .mat file has its first and third (spatial) axes swapped relative to the coordinate frame the file’s own position variables (e.g. pos_p) were written in. Enable this to build the mesh in that same, un-swapped coordinate frame, matching position data imported with swap_axes=False.

• name (str) – Name assigned to the resulting mesh. In per_label mode the label is appended as "{name}_{label}". By default "object".

Returns:

For a single mesh, a Mesh whose vertices is an (N, 3) float array and faces an (M, 4) integer array of quad vertex indices. With per_label=True, a {label: Mesh} dict.

Return type:

Mesh | dict[int, Mesh]