`spateo.tdr.models`#

Subpackages#

Package Contents#

Functions#

`alpha_shape_mesh`(→ pyvista.PolyData)	Computes a triangle mesh from a point cloud based on the alpha shape algorithm.
`ball_pivoting_mesh`(pc[, radii])	Computes a triangle mesh from an oriented point cloud based on the ball pivoting algorithm.
`construct_cells`(pc, cell_size[, geometry, xyz_scale, ...])	Reconstructing cells from point clouds.
`construct_surface`(→ Tuple[pyvista.PolyData, ...)	Surface mesh reconstruction based on 3D point cloud model.
`fix_mesh`(→ pyvista.PolyData)	Repair the mesh where it was extracted and subtle holes along complex parts of the mesh.
`marching_cube_mesh`(pc[, levelset, mc_scale_factor])	Computes a triangle mesh from a point cloud based on the marching cube algorithm.
`poisson_mesh`(→ pyvista.PolyData)	Computes a triangle mesh from an oriented point cloud based on the screened poisson reconstruction.
`pv_mesh`(→ pyvista.PolyData)	Generate a 3D tetrahedral mesh from a scattered points and extract surface mesh of the 3D tetrahedral mesh.
`uniform_larger_pc`(→ pyvista.PolyData)	Generates a uniform point cloud with a larger number of points.
`uniform_mesh`(→ pyvista.PolyData)	Generate a uniformly meshed surface using voronoi clustering.
`construct_align_lines`(→ pyvista.PolyData)	Construct alignment lines between models after model alignment.
`construct_arrow`(→ pyvista.PolyData)	Create a 3D arrow model.
`construct_arrows`(→ pyvista.PolyData)	Create multiple 3D arrows model.
`construct_axis_line`(→ pyvista.PolyData)	Construct axis line.
`construct_bounding_box`(, grid_num, ...)	Construct a bounding box model of the model.
`construct_field`(→ pyvista.PolyData)	Create a 3D vector field arrows model.
`construct_field_streams`(model[, vf_key, ...])	Integrate a vector field to generate streamlines.
`construct_genesis`(→ pyvista.MultiBlock)	Reconstruction of cell-level cell developmental change model based on the cell fate prediction results. Here we only
`construct_genesis_X`(→ pyvista.MultiBlock)	Reconstruction of cell-level cell developmental change model based on the cell fate prediction results. Here we only
`construct_line`(→ pyvista.PolyData)	Create a 3D line model.
`construct_lines`(→ pyvista.PolyData)	Create 3D lines model.
`construct_space`(, grid_num, Tuple[int]]] = None, ...)	Construct a model(space-model) with uniform spacing in the three coordinate directions.
`construct_trajectory`(→ pyvista.PolyData)	Reconstruction of cell developmental trajectory model based on cell fate prediction.
`construct_trajectory_X`(→ pyvista.PolyData)	Reconstruction of cell developmental trajectory model.
`construct_pc`(→ pyvista.PolyData)	Construct a point cloud model based on 3D coordinate information.
`add_model_labels`(→ PolyData or UnstructuredGrid)	Add rgba color to each point of model based on labels.
`center_to_zero`(model[, inplace])	Translate the center point of the model to the (0, 0, 0).
`collect_models`(→ pyvista.MultiBlock)	A composite class to hold many data sets which can be iterated over.
`merge_models`(→ PolyData or UnstructuredGrid)	Merge all models in the models list. The format of all models must be the same.
`multiblock2model`(model[, message])	Merge all models in MultiBlock into one model
`read_model`(filename)	Read any file type supported by vtk or meshio.
`rotate_model`(, rotate_center, tuple] = None, inplace, ...)	Rotate the model around the rotate_center.
`save_model`(model, filename[, binary, texture])	Save the pvvista/vtk model to vtk/vtm file.
`scale_model`(→ Union[pyvista.PolyData, ...)	Scale the model around the center of the model.
`translate_model`(, inplace, pyvista.UnstructuredGrid, None])	Translate the mesh.
`voxelize_mesh`(→ pyvista.UnstructuredGrid)	Construct a volumetric mesh based on surface mesh.
`voxelize_pc`(→ pyvista.UnstructuredGrid)	Voxelize the point cloud.

spateo.tdr.models.alpha_shape_mesh(pc: pyvista.PolyData, alpha: float = 2.0) → pyvista.PolyData#

Computes a triangle mesh from a point cloud based on the alpha shape algorithm. Algorithm Overview:

For each real number α, define the concept of a generalized disk of radius 1/α as follows:

If α = 0, it is a closed half-plane; If α > 0, it is a closed disk of radius 1/α; If α < 0, it is the closure of the complement of a disk of radius −1/α.

Then an edge of the alpha-shape is drawn between two members of the finite point set whenever there exists a generalized disk of radius 1/α containing none of the point set and which has the property that the two points lie on its boundary. If α = 0, then the alpha-shape associated with the finite point set is its ordinary convex hull.

Parameters

pc: A point cloud model.
alpha: Parameter to control the shape. With decreasing alpha value the shape shrinks and creates cavities. A very big value will give a shape close to the convex hull.

Returns

A mesh model.

spateo.tdr.models.ball_pivoting_mesh(pc: pyvista.PolyData, radii: List[float] = None)#

Computes a triangle mesh from an oriented point cloud based on the ball pivoting algorithm. Algorithm Overview:

The main assumption this algorithm is based on is the following: Given three vertices, and a ball of radius r, the three vertices form a triangle if the ball is getting “caught” and settle between the points, without containing any other point. The algorithm stimulates a virtual ball of radius r. Each iteration consists of two steps:

Seed triangle - The ball rolls over the point cloud until it gets “caught” between three vertices and
settles between in them. Choosing the right r promises no other point is contained in the formed triangle. This triangle is called “Seed triangle”.

Expanding triangle - The ball pivots from each edge in the seed triangle, looking for a third point. It
pivots until it gets “caught” in the triangle formed by the edge and the third point. A new triangle is formed, and the algorithm tries to expand from it. This process continues until the ball can’t find any point to expand to.

At this point, the algorithm looks for a new seed triangle, and the process described above starts all over.

Useful Notes:

The point cloud is “dense enough”;
The chosen r size should be “slightly” larger than the average space between points.

Parameters

pc: A point cloud model.
radii: The radii of the ball that are used for the surface reconstruction. This is a list of multiple radii that will create multiple balls of different radii at the same time.

Returns

A mesh model.

spateo.tdr.models.construct_cells(pc: pyvista.PolyData, cell_size: numpy.ndarray, geometry: Literal[cube, sphere, ellipsoid] = 'cube', xyz_scale: tuple = (1, 1, 1), n_scale: tuple = (1, 1), factor: float = 0.5)#

Reconstructing cells from point clouds.

Parameters

pc

A point cloud object, including pc.point_data["obs_index"].

geometry

The geometry of generating cells. Available geometry are:

geometry = 'cube'
geometry = 'sphere'
geometry = 'ellipsoid'

cell_size

A numpy.ndarray object including the relative radius/length size of each cell.

xyz_scale

The scale factor for the x-axis, y-axis and z-axis.

n_scale

The squareness parameter in the x-y plane adn z axis. Only works if geometry = 'ellipsoid'.

factor

Scale factor applied to scaling array.

Returns

A cells mesh including ds_glyph.point_data[“cell_size”], ds_glyph.point_data[“cell_centroid”] and the data contained in the pc.

Return type

ds_glyph

spateo.tdr.models.construct_surface(pc: pyvista.PolyData, key_added: str = 'groups', label: str = 'surface', color: Optional[str] = 'gainsboro', alpha: Union[float, int] = 1.0, uniform_pc: bool = False, uniform_pc_alpha: Union[float, int] = 0, cs_method: Literal[pyvista, alpha_shape, ball_pivoting, poisson, marching_cube] = 'marching_cube', cs_args: Optional[dict] = None, nsub: Optional[int] = 3, nclus: int = 20000, smooth: Optional[int] = 1000, scale_distance: Union[float, int, list, tuple] = None, scale_factor: Union[float, int, list, tuple] = None) → Tuple[pyvista.PolyData, pyvista.PolyData]#

Surface mesh reconstruction based on 3D point cloud model.

Parameters

pc

A point cloud model.

key_added

The key under which to add the labels.

label

The label of reconstructed surface mesh model.

color

Color to use for plotting mesh. The default color is 'gainsboro'.

alpha

The opacity of the color to use for plotting mesh. The default alpha is 0.8.

uniform_pc

Generates a uniform point cloud with a larger number of points.

uniform_pc_alpha

Specify alpha (or distance) value to control output of this filter.

cs_method

The methods of generating a surface mesh. Available cs_method are:

'pyvista': Generate a 3D tetrahedral mesh based on pyvista.
'alpha_shape': Computes a triangle mesh on the alpha shape algorithm.
'ball_pivoting': Computes a triangle mesh based on the Ball Pivoting algorithm.
'poisson': Computes a triangle mesh based on thee Screened Poisson Reconstruction.
'marching_cube': Computes a triangle mesh based on the marching cube algorithm.

cs_args

Parameters for various surface reconstruction methods. Available cs_args are: * 'pyvista': {‘alpha’: 0} * 'alpha_shape': {‘alpha’: 2.0} * 'ball_pivoting': {‘radii’: [1]} * 'poisson': {‘depth’: 8, ‘width’=0, ‘scale’=1.1, ‘linear_fit’: False, ‘density_threshold’: 0.01} * 'marching_cube': {‘levelset’: 0, ‘mc_scale_factor’: 1}

nsub

Number of subdivisions. Each subdivision creates 4 new triangles, so the number of resulting triangles is nface*4**nsub where nface is the current number of faces.

nclus

Number of voronoi clustering.

smooth

Number of iterations for Laplacian smoothing.

scale_distance

The distance by which the model is scaled. If scale_distance is float, the model is scaled same distance along the xyz axis; when the scale factor is list, the model is scaled along the xyz axis at different distance. If scale_distance is None, there will be no scaling based on distance.

scale_factor

The scale by which the model is scaled. If scale factor is float, the model is scaled along the xyz axis at the same scale; when the scale factor is list, the model is scaled along the xyz axis at different scales. If scale_factor is None, there will be no scaling based on scale factor.

Returns

A reconstructed surface mesh, which contains the following properties:: uniform_surf.cell_data[key_added], the label array; uniform_surf.cell_data[f'{key_added}_rgba'], the rgba colors of the label array.
inside_pc: A point cloud, which contains the following properties:: inside_pc.point_data['obs_index'], the obs_index of each coordinate in the original adata. inside_pc.point_data[key_added], the groupby information. inside_pc.point_data[f'{key_added}_rgba'], the rgba colors of the groupby information.

Return type

uniform_surf

spateo.tdr.models.fix_mesh(mesh: pyvista.PolyData) → pyvista.PolyData#: Repair the mesh where it was extracted and subtle holes along complex parts of the mesh.

spateo.tdr.models.marching_cube_mesh(pc: pyvista.PolyData, levelset: Union[int, float] = 0, mc_scale_factor: Union[int, float] = 1.0)#

Computes a triangle mesh from a point cloud based on the marching cube algorithm. Algorithm Overview:

The algorithm proceeds through the scalar field, taking eight neighbor locations at a time (thus forming an imaginary cube), then determining the polygon(s) needed to represent the part of the iso-surface that passes through this cube. The individual polygons are then fused into the desired surface.

Parameters

pc: A point cloud model.
levelset: The levelset of iso-surface. It is recommended to set levelset to 0 or 0.5.
mc_scale_factor: The scale of the model. The scaled model is used to construct the mesh model.

Returns

A mesh model.

spateo.tdr.models.poisson_mesh(pc: pyvista.PolyData, depth: int = 8, width: float = 0, scale: float = 1.1, linear_fit: bool = False, density_threshold: Optional[float] = None) → pyvista.PolyData#

Computes a triangle mesh from an oriented point cloud based on the screened poisson reconstruction.

Parameters

pc

A point cloud model.

depth

Maximum depth of the tree that will be used for surface reconstruction. Running at depth d corresponds to solving on a grid whose resolution is no larger than 2^d x 2^d x 2^d.

Note that since the reconstructor adapts the octree to the sampling density, the specified reconstruction depth is only an upper bound.

The depth that defines the depth of the octree used for the surface reconstruction and hence implies the resolution of the resulting triangle mesh. A higher depth value means a mesh with more details.

width

Specifies the target width of the finest level octree cells. This parameter is ignored if depth is specified.

scale

Specifies the ratio between the diameter of the cube used for reconstruction and the diameter of the samples’ bounding cube.

linear_fit

If true, the reconstructor will use linear interpolation to estimate the positions of iso-vertices.

density_threshold

The threshold of the low density.

Returns

A mesh model.

spateo.tdr.models.pv_mesh(pc: pyvista.PolyData, alpha: float = 2.0) → pyvista.PolyData#

Generate a 3D tetrahedral mesh from a scattered points and extract surface mesh of the 3D tetrahedral mesh.

Parameters

pc: A point cloud model.
alpha: Distance value to control output of this filter. For a non-zero alpha value, only vertices, edges, faces, or tetrahedron contained within the circumspect (of radius alpha) will be output. Otherwise, only tetrahedron will be output.

Returns

A mesh model.

spateo.tdr.models.uniform_larger_pc(pc: pyvista.PolyData, alpha: Union[float, int] = 0, nsub: Optional[int] = 5, nclus: int = 20000) → pyvista.PolyData#

Generates a uniform point cloud with a larger number of points. If the number of points in the original point cloud is too small or the distribution of the original point cloud is not uniform, making it difficult to construct the surface, this method can be used for preprocessing.

Parameters

pc: A point cloud model.
alpha: Specify alpha (or distance) value to control output of this filter. For a non-zero alpha value, only edges or triangles contained within a sphere centered at mesh vertices will be output. Otherwise, only triangles will be output.
nsub: Number of subdivisions. Each subdivision creates 4 new triangles, so the number of resulting triangles is nface*4**nsub where nface is the current number of faces.
nclus: Number of voronoi clustering.

Returns

A uniform point cloud with a larger number of points.

Return type

new_pc

spateo.tdr.models.uniform_mesh(mesh: pyvista.PolyData, nsub: Optional[int] = 3, nclus: int = 20000) → pyvista.PolyData#

Generate a uniformly meshed surface using voronoi clustering.

Parameters

mesh: A mesh model.
nsub: Number of subdivisions. Each subdivision creates 4 new triangles, so the number of resulting triangles is nface*4**nsub where nface is the current number of faces.
nclus: Number of voronoi clustering.

Returns

A uniform mesh model.

Return type

new_mesh

spateo.tdr.models.construct_align_lines(model1_points: numpy.ndarray, model2_points: numpy.ndarray, key_added: str = 'check_alignment', label: Union[str, list, numpy.ndarray] = 'align_mapping', color: Union[str, list, dict, numpy.ndarray] = 'gainsboro', alpha: Union[float, int, list, dict, numpy.ndarray] = 1.0) → pyvista.PolyData#

Construct alignment lines between models after model alignment.

Parameters

model1_points: Start location in model1 of the line.
model2_points: End location in model2 of the line.
key_added: The key under which to add the labels.
label: The label of alignment lines model.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.

Returns

Alignment lines model.

spateo.tdr.models.construct_arrow(start_point: Union[list, tuple, numpy.ndarray], direction: Union[list, tuple, numpy.ndarray], arrow_scale: Optional[Union[int, float]] = None, key_added: Optional[str] = 'arrow', label: str = 'arrow', color: str = 'gainsboro', alpha: float = 1.0, **kwargs) → pyvista.PolyData#

Create a 3D arrow model.

Parameters

start_point: Start location in [x, y, z] of the arrow.
direction: Direction the arrow points to in [x, y, z].
arrow_scale: Scale factor of the entire object. ‘auto’ scales to length of direction array.
key_added: The key under which to add the labels.
label: The label of arrow model.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.
**kwargs: Additional parameters that will be passed to _construct_arrow function.

Returns

Arrow model.

spateo.tdr.models.construct_arrows(start_points: numpy.ndarray, direction: numpy.ndarray = None, arrows_scale: Optional[numpy.ndarray] = None, n_sampling: Optional[int] = None, sampling_method: str = 'trn', factor: float = 1.0, key_added: Optional[str] = 'arrow', label: Union[str, list, numpy.ndarray] = 'arrows', color: Union[str, list, dict, numpy.ndarray] = 'gainsboro', alpha: Union[float, int, list, dict, numpy.ndarray] = 1.0, **kwargs) → pyvista.PolyData#

Create multiple 3D arrows model.

Parameters

start_points: List of Start location in [x, y, z] of the arrows.
direction: Direction the arrows points to in [x, y, z].
arrows_scale: Scale factor of the entire object.
n_sampling: n_sampling is the number of coordinates to keep after sampling. If there are too many coordinates in start_points, the generated arrows model will be too complex and unsightly, so sampling is used to reduce the number of coordinates.
sampling_method: The method to sample data points, can be one of ['trn', 'kmeans', 'random'].
factor: Scale factor applied to scaling array.
key_added: The key under which to add the labels.
label: The label of arrows models.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.
**kwargs: Additional parameters that will be passed to _construct_arrow function.

Returns

Arrows model.

spateo.tdr.models.construct_axis_line(axis_points: numpy.ndarray, key_added: str = 'axis', label: str = 'axis_line', color: str = 'gainsboro', alpha: Union[float, int, list, dict, numpy.ndarray] = 1.0) → pyvista.PolyData#

Construct axis line.

Parameters

axis_points: List of points defining an axis.
key_added: The key under which to add the labels.
label: The label of axis line model.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.

Returns

Axis line model.

spateo.tdr.models.construct_bounding_box(model: Union[pyvista.DataSet, pyvista.MultiBlock], expand_dist: Union[int, float, list, tuple] = (0, 0, 0), grid_num: Optional[Union[List[int], Tuple[int]]] = None, key_added: str = 'bounding_box', label: str = 'bounding_box', color: str = 'gainsboro', alpha: float = 0.5) → pyvista.PolyData#

Construct a bounding box model of the model.

Parameters

model: A three dims model.
expand_dist: The length of space-model to be extended in all directions.
grid_num: Number of grid to generate.
key_added: The key under which to add the labels.
label: The label of space-model.
color: Color to use for plotting space-model.
alpha: The opacity of the color to use for plotting space-model.

Returns

A bounding box model.

spateo.tdr.models.construct_field(model: pyvista.PolyData, vf_key: str = 'VecFld_morpho', arrows_scale_key: Optional[str] = None, n_sampling: Optional[int] = None, sampling_method: str = 'trn', factor: float = 1.0, key_added: str = 'v_arrows', label: Union[str, list, numpy.ndarray] = 'vector field', color: Union[str, list, dict, numpy.ndarray] = 'gainsboro', alpha: float = 1.0, **kwargs) → pyvista.PolyData#

Create a 3D vector field arrows model.

Parameters

model: A model that provides coordinate information and vector information for constructing vector field models.
vf_key: The key under which are the vector information.
arrows_scale_key: The key under which are scale factor of the entire object.
n_sampling: n_sampling is the number of coordinates to keep after sampling. If there are too many coordinates in start_points, the generated arrows model will be too complex and unsightly, so sampling is used to reduce the number of coordinates.
sampling_method: The method to sample data points, can be one of ['trn', 'kmeans', 'random'].
factor: Scale factor applied to scaling array.
key_added: The key under which to add the labels.
label: The label of arrows models.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.
**kwargs: Additional parameters that will be passed to construct_arrows function.

Returns

A 3D vector field arrows model.

spateo.tdr.models.construct_field_streams(model: pyvista.PolyData, vf_key: str = 'VecFld_morpho', source_center: Optional[Tuple[float]] = None, source_radius: Optional[float] = None, tip_factor: Union[int, float] = 10, tip_radius: float = 0.2, key_added: str = 'v_streams', label: Union[str, list, numpy.ndarray] = 'vector field', stream_color: str = 'gainsboro', tip_color: str = 'orangered', alpha: float = 1.0, **kwargs)#

Integrate a vector field to generate streamlines.

Parameters

model: A model that provides coordinate information and vector information for constructing vector field models.
vf_key: The key under which are the active vector field information.
source_center: Length 3 tuple of floats defining the center of the source particles. Defaults to the center of the dataset.
source_radius: Float radius of the source particle cloud. Defaults to one-tenth of the diagonal of the dataset’s spatial extent.
tip_factor: Scale factor applied to scaling the tips.
tip_radius: Radius of the tips.
key_added: The key under which to add the labels.
label: The label of arrows models.
stream_color: Color to use for plotting streamlines.
tip_color: Color to use for plotting tips.
alpha: The opacity of the color to use for plotting model.
**kwargs: Additional parameters that will be passed to streamlines function.

Returns

3D vector field streamlines model. src: The source particles as pyvista.PolyData as well as the streamlines.

Return type

streams_model

spateo.tdr.models.construct_genesis(adata: anndata.AnnData, fate_key: str = 'fate_morpho', n_steps: int = 100, logspace: bool = False, t_end: Optional[Union[int, float]] = None, key_added: str = 'genesis', label: Optional[Union[str, list, numpy.ndarray]] = None, color: Union[str, list, dict] = 'skyblue', alpha: Union[float, list, dict] = 1.0) → pyvista.MultiBlock#

Reconstruction of cell-level cell developmental change model based on the cell fate prediction results. Here we only need to enter the three-dimensional coordinates of the cells at different developmental stages.

Parameters

adata: AnnData object that contains the fate prediction in the .uns attribute.
fate_key: The key under which are the active fate information.
n_steps: The number of times steps fate prediction will take.
logspace: Whether or to sample time points linearly on log space. If not, the sorted unique set of all times points from all cell states’ fate prediction will be used and then evenly sampled up to n_steps time points.
t_end: The length of the time period from which to predict cell state forward or backward over time.
key_added: The key under which to add the labels.
label: The label of cell developmental change model. If label == None, the label will be automatically generated.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.

Returns

A MultiBlock contains cell models for all stages.

spateo.tdr.models.construct_genesis_X(stages_X: List[numpy.ndarray], n_spacing: Optional[int] = None, key_added: str = 'genesis', label: Optional[Union[str, list, numpy.ndarray]] = None, color: Union[str, list, dict] = 'skyblue', alpha: Union[float, list, dict] = 1.0) → pyvista.MultiBlock#

Parameters

stages_X: The three-dimensional coordinates of the cells at different developmental stages.
n_spacing: Subdivided into n_spacing time points between two periods.
key_added: The key under which to add the labels.
label: The label of cell developmental change model. If label == None, the label will be automatically generated.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.

Returns

A MultiBlock contains cell models for all stages.

spateo.tdr.models.construct_line(start_point: Union[list, tuple, numpy.ndarray], end_point: Union[list, tuple, numpy.ndarray], key_added: Optional[str] = 'line', label: str = 'line', color: str = 'gainsboro', alpha: float = 1.0) → pyvista.PolyData#

Create a 3D line model.

Parameters

start_point: Start location in [x, y, z] of the line.
end_point: End location in [x, y, z] of the line.
key_added: The key under which to add the labels.
label: The label of line model.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.

Returns

Line model.

spateo.tdr.models.construct_lines(points: numpy.ndarray, edges: numpy.ndarray, key_added: Optional[str] = 'line', label: Union[str, list, numpy.ndarray] = 'lines', color: Union[str, list, dict] = 'gainsboro', alpha: Union[float, int, list, dict] = 1.0) → pyvista.PolyData#

Create 3D lines model.

Parameters

points: List of points.
edges: The edges between points.
key_added: The key under which to add the labels.
label: The label of lines model.
color: Color to use for plotting model.
alpha: The opacity of the color to use for plotting model.

Returns

Lines model.

spateo.tdr.models.construct_space(model: Union[pyvista.DataSet, pyvista.MultiBlock], expand_dist: Union[int, float, list, tuple] = (0, 0, 0), grid_num: Optional[Union[List[int], Tuple[int]]] = None, key_added: Optional[str] = 'space', label: str = 'space', color: str = 'gainsboro', alpha: float = 0.5) → pyvista.UniformGrid#

Construct a model(space-model) with uniform spacing in the three coordinate directions. The six surfaces of the commonly generated space-model are exactly the boundaries of the model, but the space-model can also be expanded by expand_dist.

Parameters

model: A three dims model.
expand_dist: The length of space-model to be extended in all directions.
grid_num: Number of grid to generate.
key_added: The key under which to add the labels.
label: The label of space-model.
color: Color to use for plotting space-model.
alpha: The opacity of the color to use for plotting space-model.

Returns

A space-model with uniform spacing in the three coordinate directions.

spateo.tdr.models.construct_trajectory(adata: anndata.AnnData, fate_key: str = 'fate_develop', n_sampling: Optional[int] = None, sampling_method: str = 'trn', key_added: str = 'trajectory', label: Optional[Union[str, list, numpy.ndarray]] = None, tip_factor: Union[int, float] = 5, tip_radius: float = 0.2, trajectory_color: Union[str, list, dict] = 'gainsboro', tip_color: Union[str, list, dict] = 'orangered', alpha: float = 1.0) → pyvista.PolyData#

Reconstruction of cell developmental trajectory model based on cell fate prediction.

Parameters

adata: AnnData object that contains the fate prediction in the .uns attribute.
fate_key: The key under which are the active fate information.
n_sampling: n_sampling is the number of coordinates to keep after sampling. If there are too many coordinates in start_points, the generated arrows model will be too complex and unsightly, so sampling is used to reduce the number of coordinates.
sampling_method: The method to sample data points, can be one of ['trn', 'kmeans', 'random'].
key_added: The key under which to add the labels.
label: The label of trajectory model.
tip_factor: Scale factor applied to scaling the tips.
tip_radius: Radius of the tips.
trajectory_color: Color to use for plotting trajectory model.
tip_color: Color to use for plotting tips.
alpha: The opacity of the color to use for plotting model.

Returns

3D cell developmental trajectory model.

Return type

trajectory_model

spateo.tdr.models.construct_trajectory_X(cells_states: Union[numpy.ndarray, List[numpy.ndarray]], init_states: Optional[numpy.ndarray] = None, n_sampling: Optional[int] = None, sampling_method: str = 'trn', key_added: str = 'trajectory', label: Optional[Union[str, list, numpy.ndarray]] = None, tip_factor: Union[int, float] = 5, tip_radius: float = 0.2, trajectory_color: Union[str, list, dict] = 'gainsboro', tip_color: Union[str, list, dict] = 'orangered', alpha: Union[float, list, dict] = 1.0) → pyvista.PolyData#

Reconstruction of cell developmental trajectory model.

Parameters

cells_states: Three-dimensional coordinates of all cells at all times points.
init_states: Three-dimensional coordinates of all cells at the starting time point.
n_sampling: n_sampling is the number of coordinates to keep after sampling. If there are too many coordinates in start_points, the generated arrows model will be too complex and unsightly, so sampling is used to reduce the number of coordinates.
sampling_method: The method to sample data points, can be one of ['trn', 'kmeans', 'random'].
key_added: The key under which to add the labels.
label: The label of trajectory model.
tip_factor: Scale factor applied to scaling the tips.
tip_radius: Radius of the tips.
trajectory_color: Color to use for plotting trajectory model.
tip_color: Color to use for plotting tips.
alpha: The opacity of the color to use for plotting model.

Returns

3D cell developmental trajectory model.

Return type

trajectory_model

spateo.tdr.models.construct_pc(adata: anndata.AnnData, spatial_key: str = 'spatial', groupby: Union[str, tuple] = None, key_added: str = 'groups', mask: Union[str, int, float, list] = None, colormap: Union[str, list, dict] = 'rainbow', alphamap: Union[float, list, dict] = 1.0) → pyvista.PolyData#

Construct a point cloud model based on 3D coordinate information.

Parameters

adata: AnnData object.
spatial_key: The key in .obsm that corresponds to the spatial coordinate of each bucket.
groupby: The key that stores clustering or annotation information in .obs, a gene name or a list of gene names in .var.
key_added: The key under which to add the labels.
mask: The part that you don’t want to be displayed.
colormap: Colors to use for plotting pcd. The default colormap is 'rainbow'.
alphamap: The opacity of the colors to use for plotting pcd. The default alphamap is 1.0.

Returns

A point cloud, which contains the following properties:: pc.point_data[key_added], the groupby information. pc.point_data[f'{key_added}_rgba'], the rgba colors of the groupby information. pc.point_data['obs_index'], the obs_index of each coordinate in the original adata.

Return type

spateo.tdr.models.add_model_labels(model: Union[pyvista.PolyData, pyvista.UnstructuredGrid, pyvista.UniformGrid], labels: numpy.ndarray, key_added: str = 'groups', where: Literal[point_data, cell_data] = 'cell_data', colormap: Union[str, list, dict, numpy.ndarray] = 'rainbow', alphamap: Union[float, list, dict, numpy.ndarray] = 1.0, mask_color: Optional[str] = 'gainsboro', mask_alpha: Optional[float] = 0.0, inplace: bool = False) → PolyData or UnstructuredGrid#

Add rgba color to each point of model based on labels.

Parameters

model: A reconstructed model.
labels: An array of labels of interest.
key_added: The key under which to add the labels.
where: The location where the label information is recorded in the model.
colormap: Colors to use for plotting data.
alphamap: The opacity of the color to use for plotting data.
mask_color: Color to use for plotting mask information.
mask_alpha: The opacity of the color to use for plotting mask information.
inplace: Updates model in-place.

Returns

model.cell_data[key_added] or model.point_data[key_added], the labels array; model.cell_data[f'{key_added}_rgba'] or model.point_data[f'{key_added}_rgba'], the rgba colors of the labels.

Return type

A model, which contains the following properties

spateo.tdr.models.center_to_zero(model: Union[pyvista.PolyData, pyvista.UnstructuredGrid], inplace: bool = False)#

Translate the center point of the model to the (0, 0, 0).

Parameters

model: A 3D reconstructed model.
inplace: Updates model in-place.

Returns

Model with center point at (0, 0, 0).

Return type

model_z

spateo.tdr.models.collect_models(models: List[PolyData or UnstructuredGrid or DataSet], models_name: Optional[List[str]] = None) → pyvista.MultiBlock#

A composite class to hold many data sets which can be iterated over. You can think of MultiBlock like lists or dictionaries as we can iterate over this data structure by index and we can also access blocks by their string name. If the input is a dictionary, it can be iterated in the following ways:

>>> blocks = collect_models(models, models_name)
>>> for name in blocks.keys():
...     print(blocks[name])

If the input is a list, it can be iterated in the following ways:

>>> blocks = collect_models(models)
>>> for block in blocks:
...    print(block)

spateo.tdr.models.merge_models(models: List[PolyData or UnstructuredGrid or DataSet]) → PolyData or UnstructuredGrid#: Merge all models in the models list. The format of all models must be the same.

spateo.tdr.models.multiblock2model(model, message=None)#: Merge all models in MultiBlock into one model

spateo.tdr.models.read_model(filename: str)#

Read any file type supported by vtk or meshio. :param filename: The string path to the file to read.

Returns: Wrapped PyVista dataset.

spateo.tdr.models.rotate_model(model: Union[pyvista.PolyData, pyvista.UnstructuredGrid], angle: Union[list, tuple] = (0, 0, 0), rotate_center: Union[list, tuple] = None, inplace: bool = False) → Union[pyvista.PolyData, pyvista.UnstructuredGrid, None]#

Rotate the model around the rotate_center.

Parameters

model: A 3D reconstructed model.
angle: Angles in degrees to rotate about the x-axis, y-axis, z-axis. Length 3 list or tuple.
rotate_center: Rotation center point. The default is the center of the model. Length 3 list or tuple.
inplace: Updates model in-place.

Returns

The rotated model.

Return type

model_r

spateo.tdr.models.save_model(model: Union[pyvista.DataSet, pyvista.MultiBlock], filename: str, binary: bool = True, texture: Union[str, numpy.ndarray] = None)#

Save the pvvista/vtk model to vtk/vtm file. :param model: A reconstructed model. :param filename: Filename of output file. Writer type is inferred from the extension of the filename.

If model is a pyvista.MultiBlock object, please enter a filename ending with .vtm; else please enter a filename ending with .vtk.

Parameters

binary

If True, write as binary. Otherwise, write as ASCII. Binary files write much faster than ASCII and have a smaller file size.

texture

Write a single texture array to file when using a PLY file.

Texture array must be a 3 or 4 component array with the datatype np.uint8. Array may be a cell array or a point array, and may also be a string if the array already exists in the PolyData.

If a string is provided, the texture array will be saved to disk as that name. If an array is provided, the texture array will be saved as ‘RGBA’

spateo.tdr.models.scale_model(model: Union[pyvista.PolyData, pyvista.UnstructuredGrid], distance: Union[float, int, list, tuple] = None, scale_factor: Union[float, int, list, tuple] = 1, scale_center: Union[list, tuple] = None, inplace: bool = False) → Union[pyvista.PolyData, pyvista.UnstructuredGrid, None]#

Scale the model around the center of the model.

Parameters

model: A 3D reconstructed model.
distance: The distance by which the model is scaled. If distance is float, the model is scaled same distance along the xyz axis; when the scale factor is list, the model is scaled along the xyz axis at different distance. If distance is None, there will be no scaling based on distance.
scale_factor: The scale by which the model is scaled. If scale factor is float, the model is scaled along the xyz axis at the same scale; when the scale factor is list, the model is scaled along the xyz axis at different scales. If scale_factor is None, there will be no scaling based on scale factor.
scale_center: Scaling center. If scale factor is None, the scale_center will default to the center of the model.
inplace: Updates model in-place.

Returns

The scaled model.

Return type

model_s

spateo.tdr.models.translate_model(model: Union[pyvista.PolyData, pyvista.UnstructuredGrid], distance: Union[list, tuple] = (0, 0, 0), inplace: bool = False) → Union[pyvista.PolyData, pyvista.UnstructuredGrid, None]#

Translate the mesh.

Parameters

model: A 3D reconstructed model.
distance: Distance to translate about the x-axis, y-axis, z-axis. Length 3 list or tuple.
inplace: Updates model in-place.

Returns

The translated model.

Return type

model_t

spateo.tdr.models.voxelize_mesh(mesh: Union[pyvista.PolyData, pyvista.UnstructuredGrid], voxel_pc: Union[pyvista.PolyData, pyvista.UnstructuredGrid] = None, key_added: str = 'groups', label: str = 'voxel', color: Optional[str] = 'gainsboro', alpha: Union[float, int] = 1.0, smooth: Optional[int] = 200) → pyvista.UnstructuredGrid#

Construct a volumetric mesh based on surface mesh.

Parameters

mesh: A surface mesh model.
voxel_pc: A voxel model which contains the voxel_pc.cell_data['obs_index'] and voxel_pc.cell_data[key_added].
key_added: The key under which to add the labels.
label: The label of reconstructed voxel model.
color: Color to use for plotting mesh. The default color is 'gainsboro'.
alpha: The opacity of the color to use for plotting model. The default alpha is 0.8.
smooth: The smoothness of the voxel model.

Returns

A reconstructed voxel model, which contains the following properties:: voxel_model.cell_data[key_added], the label array; voxel_model.cell_data[f’{key_added}_rgba’], the rgba colors of the label array. voxel_model.cell_data[‘obs_index’], the cell labels if not (voxel_pc is None).

Return type

voxel_model

spateo.tdr.models.voxelize_pc(pc: pyvista.PolyData, voxel_size: Optional[numpy.ndarray] = None) → pyvista.UnstructuredGrid#

Voxelize the point cloud.

Parameters

pc: A point cloud model.
voxel_size: The size of the voxelized points. The shape of voxel_size is (pc.n_points, 3).

Returns

A voxel model.

Return type

voxel

spateo.tdr.models#

Subpackages#

Package Contents#

Functions#

`spateo.tdr.models`#