gnp package

Subpackages

• gnp.dataset module
  • gnp.dataset.patch
    • PatchGenerator
      • PatchGenerator.get_centers()
      • PatchGenerator.get_patch()
      • PatchGenerator.get_patch_indices()
    • PatchLoader
    • graph_from_dict()
  • gnp.dataset.pca
    • PCABatch
      • PCABatch.to_pca()
    • PCAPatch
      • PCAPatch.initialize()
      • PCAPatch.to_pca()
  • Module contents
• gnp.geometry module
  • gnp.geometry.surface
    • SurfacePatch
      • SurfacePatch.det_metric
      • SurfacePatch.gaussian_curvature
      • SurfacePatch.inverse_metric
      • SurfacePatch.inverse_metric_derivatives
      • SurfacePatch.laplace_beltrami_first_terms
      • SurfacePatch.laplace_beltrami_from_coefficients()
      • SurfacePatch.laplace_beltrami_legendre_basis
      • SurfacePatch.laplace_beltrami_second_terms
      • SurfacePatch.local_coordinate_basis
      • SurfacePatch.mean_curvature
      • SurfacePatch.metric
      • SurfacePatch.normals
      • SurfacePatch.normals_pca
      • SurfacePatch.pca_coordinates
      • SurfacePatch.shape
      • SurfacePatch.tangents
      • SurfacePatch.tangents_pca
      • SurfacePatch.weingarten
      • SurfacePatch.xyz_coordinates
  • gnp.geometry.legendre
    • Legendre1D
      • Legendre1D.evaluate()
      • Legendre1D.first_derivative()
      • Legendre1D.second_derivative()
    • Legendre2D
      • Legendre2D.batch_ls_best_fit()
      • Legendre2D.compute_from_coeffs()
      • Legendre2D.derivatives_from_coeffs()
      • Legendre2D.evaluate()
      • Legendre2D.evaluate_derivatives()
      • Legendre2D.evaluate_from_coeffs()
  • Module contents
• gnp.models module
  • gnp.models.gnp
    • BlockGNP
      • BlockGNP.forward()
    • PatchGNP
      • PatchGNP.forward()
  • gnp.models.layers
    • BlockConv
      • BlockConv.forward()
      • BlockConv.message()
      • BlockConv.update()
    • BlockKernel
      • BlockKernel.forward()
  • Module contents

Submodules

gnp.config module

gnp.config.load_config(path: Path) → dict

Load a configuration file from a yaml file.

Parameters:

path (str) -- Path to the yaml file.

Returns:

Dictionary containing the configuration parameters.

Return type:

dict

gnp.config.load_model(model_path: Path) → PatchGNP

Load a model from a directory.

Parameters:

model_dir (Path) -- Path to the model directory.

Returns:

The loaded model.

Return type:

torch.nn.Module

gnp.config.load_patchloader(cfg: dict, **kwargs) → PatchLoader

Load a PatchLoader using a yaml file in data_dir.

Parameters:

cfg (dict) -- Dictionary containing the configuration parameters.

gnp.estimator module

class gnp.estimator.GeometryEstimator(pcd: Tensor, orientation: Tensor, function_values: Tensor | None = None, model: str = 'clean_30k', device: device = 'cpu', **patch_kwargs: dict | None)

Bases: object

Geometry estimator for a point cloud.

Parameters:

• pcd (torch.Tensor) -- Point cloud data.

• orientation (torch.Tensor) -- Orientation data.

• function_values (Optional[torch.Tensor]) -- Function values for the point cloud.

• model (nn.Module) -- Model name.

• device (torch.device) -- Device to run the model on.

• data (dict) -- Dictionary containing the input data.

• model_path (Path) -- Path to the model state dictionary.

• model -- Loaded model.

• config_path (Path) -- Path to the configuration file.

• cfg (dict) -- Configuration dictionary.

• patch_loader (PatchLoader) -- Patch loader object.

estimate_quantities(scalar_names: list[str]) → dict

Estimate geometric quantities on the point cloud. This function returns a dictionary containing the estimated scalar and/or vector values.

Parameters:

scalar_names (list[str]) -- List of scalar names to estimate. This can be any of the following: 'xyz_coordinates', 'normals', 'tangents', 'mean_curvature', 'gaussian_curvature', 'pca_coordinates', 'normals_pca', 'tangents_pca', 'metric', 'shape', 'weingarten', 'inverse_metric', 'inverse_metric_derivatives', 'det_metric', 'laplace_beltrami_from_coefficients'

Returns:

Dictionary containing the estimated scalar values.

Return type:

dict

flow_step(delta_t: float, subsample_radius: float, smooth_radius: float, smooth_x: bool) → dict

Perform a single step of mean curvature flow on the point cloud.

Parameters:

• delta_t (float) -- Time step for the flow.

• subsample_radius (float) -- Radius used for subsampling points.

• smooth_radius (float) -- Radius used for smoothing mean curvature.

• smooth_x (bool) -- Whether to smooth the point cloud before flow.

Returns:

Dictionary containing the update point cloud data, normals, and mean curvature.

Return type:

dict

gmls_weights(batch: Batch, radius: float = 1.0, p: int = 4) → list[Tensor]

Compute the weights for the generalized moving least squares (GMLS) method.

Parameters:

• batch (Batch) -- Batch containing the input data.

• radius (float, optional) -- Radius to truncate weight function. Defaults to 1..

• p (int, optional) -- Degree p of the weight function. Defaults to 4.

Returns:

List of weight matrices for each batch.

Return type:

list[torch.Tensor]

laplace_beltrami_legendre_blocks(surface: SurfacePatch) → list[Tensor]

Laplace-Beltrami operator of Legendre basis functions at each center point.

Parameters:

surface (SurfacePatch) -- Surface patch object.

Returns:

List of outputs at each center point.

Return type:

list[torch.Tensor]

legendre_blocks(surface: SurfacePatch) → list[Tensor]

Legendre basis functions blocked by batch.

Parameters:

surface (SurfacePatch) -- Surface patch object.

Returns:

List of outputs per batch.

Return type:

list[torch.Tensor]

mean_flow(num_steps: int, save_data_per_step: int, delta_t: float, subsample_radius: float, smooth_radius: float, smooth_x: bool) → list[dict]

Perform mean curvature flow on the point cloud.

Parameters:

• num_steps (int) -- Number of steps to perform.

• save_data_per_step (int) -- Save data every n steps.

• delta_t (float) -- Time step for the flow.

• subsample_radius (float) -- Radius used for subsampling points.

• smooth_radius (float) -- Radius used for smoothing mean curvature.

• smooth_x (bool) -- Whether to smooth the point cloud before flow.

Returns:

List of dictionaries containing the updated point cloud data and normals at each saved time step.

Return type:

list[dict]

stiffness_matrix_gmls(drop_ratio: float = 0.1, radius: float = 1.0, p: int = 4, remove_outliers: bool = False) → coo_matrix

Compute the stiffness matrix using the generalized moving least squares (GMLS) method.

Parameters:

• drop_ratio (float, optional) -- Ratio of points to drop. Defaults to 0.1.

• radius (float, optional) -- Radius to truncate weight function. Defaults to 1..

• p (int, optional) -- Degree p of the weight function. Defaults to 4.

• remove_outliers (bool, optional) -- Whether to remove outliers. Defaults to False.

Returns:

Stiffness matrix.

Return type:

sp.coo_matrix

stiffness_on_batch(batch: Batch, radius: float = 1.0, p: int = 4) → tuple[Tensor, Tensor]

Compute the stiffness matrix on a batch of data.

Parameters:

• batch (Batch) -- Batch containing the input data.

• radius (float, optional) -- Radius to truncate weight function. Defaults to 1..

• p (int, optional) -- Degree p of the weight function. Defaults to 4.

Returns:

Indices and values of the stiffness matrix.

Return type:

torch.Tensor

surface_patch(patch_data: Batch) → SurfacePatch

Create a SurfacePatch from the input patch data.

Parameters:

patch_data (Batch) -- Batch containing the input patch data.

Returns:

SurfacePatch object.

Return type:

SurfacePatch

gnp.utils module

gnp.utils.smooth_values_by_gaussian(x: Tensor, values: Tensor, radius: float) → Tensor

Smooth values by a truncated Gaussian kernel over a given radius. This is used in simulating mean curvature flow.

Parameters:

• x (torch.Tensor) -- Input data points

• values (torch.Tensor) -- Values to be smoothed

• radius (float) -- Radius for truncating the Gaussian kernel. The standard deviation of the Gaussian is set to one third of the radius.

Returns:

Smoothed values

Return type:

torch.Tensor

gnp.utils.subsample_points_by_radius(x: Tensor, radius: float) → LongTensor

Subsample points by radius. This is used in simulating mean curvature flow.

Parameters:

• x (torch.Tensor) -- Points to subsample

• radius (float) -- Radius for subsampling

Returns:

Indices of subsampled points

Return type:

torch.LongTensor

Module contents