atomicds.core.utils

Functions

boxes_overlap(box1, box2)

Check if two bounding boxes overlap

convert_to_polar_coordinates(points[, ...])

Convert a set of 2D points to polar coordinates with radius and angle.

generate_graph_from_nodes(node_df)

Update a pattern graph with new node data from a DataFrame object

normalize_path(path_str)

Normalize a file path string for use with pathlib.

normalize_pixel_dimensions(points, image_shape)

Rescale pixel dimensions to a new image shape.

regions_horizontal_overlapping(node_df, ...)

Check if two regions are horizontally overlapping

rescale_cartesian_coordinates(points[, ...])

Normalize radius in polar coordinates, then convert back to cartesian to get rescaled cartesian coordinates in image dimensions.
atomicds.core.utils.normalize_pixel_dimensions(points: ndarray[tuple[Any, ...], dtype[_ScalarT]], image_shape: tuple[int, int]) → ndarray[tuple[Any, ...], dtype[_ScalarT]][source]

Rescale pixel dimensions to a new image shape.

Parameters:

  • points (NDArray) – Numpy array containing a list of points with columns width, height.

  • image_shape (tuple[int, int]) – Image shape.

Returns:

Numpy array containing the rescaled points.

Return type:

NDArray

atomicds.core.utils.boxes_overlap(box1, box2) → bool[source]

Check if two bounding boxes overlap

Parameters:

  • box1 (list[float]) – List of xmin, ymin, xmax, ymax coordinates defining first box

  • box2 (list[float]) – List of xmin, ymin, xmax, ymax coordinates defining second box

Returns:

True if the boxes overlap

Return type:

(bool)

atomicds.core.utils.regions_horizontal_overlapping(node_df: DataFrame, start_node: int, end_node: int) → bool[source]

Check if two regions are horizontally overlapping

Parameters:

  • node_df (DataFrame)

  • start_node (int)

  • end_node (int)

Return type:

bool

atomicds.core.utils.rescale_cartesian_coordinates(points: ndarray[tuple[Any, ...], dtype[_ScalarT]], origin=(0, 0), scale: float = 1.0) → ndarray[tuple[Any, ...], dtype[_ScalarT]][source]

Normalize radius in polar coordinates, then convert back to cartesian to get rescaled cartesian coordinates in image dimensions. :type points: ndarray[tuple[Any, ...], dtype[TypeVar(_ScalarT, bound= generic)]] :param points: Numpy array containing a list of points. :type points: NDArray :type origin: :param origin: Origin point. :type origin: tuple[int, int] :type scale: float :param scale: Scaling number. :type scale: float

Returns:

Numpy array containing the rescaled points.

Return type:

NDArray

atomicds.core.utils.convert_to_polar_coordinates(points: ndarray[tuple[Any, ...], dtype[_ScalarT]], origin=(0, 0), scale=1.0) → ndarray[tuple[Any, ...], dtype[_ScalarT]][source]

Convert a set of 2D points to polar coordinates with radius and angle.

Parameters:

  • points (NDArray) – Numpy array containing a list of points.

  • origin (tuple[int, int]) – Origin point.

  • scale (float) – Scaling number.

Return type:

ndarray[tuple[Any, ...], dtype[TypeVar(_ScalarT, bound= generic)]]

atomicds.core.utils.generate_graph_from_nodes(node_df: DataFrame) → Graph[source]

Update a pattern graph with new node data from a DataFrame object

Parameters:

node_df (DataFrame)

Return type:

Graph

atomicds.core.utils.normalize_path(path_str: str) → Path[source]

Normalize a file path string for use with pathlib.

This will:

  1. Remove control characters and convert “smart” quotes into plain quotes.

  2. Strip leading/trailing whitespace and any surrounding quotes.

  3. Expand user (~) and environment variables.

  4. Normalize Unicode, unify separators, and collapse “..”/“.” segments.

Parameters:

path_str (str) – Raw path string copied from Windows (may contain spaces, smart quotes, stray control chars, etc.)

Return type:

Path

Returns:

A pathlib.Path pointing to the normalized path.