dacapo.gp.elastic_augment_fuse

Module Contents

Classes

ElasticAugment

Elasticly deform a batch. Requests larger batches upstream to avoid data

Attributes

logger

dacapo.gp.elastic_augment_fuse.logger
class dacapo.gp.elastic_augment_fuse.ElasticAugment(control_point_spacing, control_point_displacement_sigma, rotation_interval, subsample=1, augmentation_probability=1.0, seed=None, uniform_3d_rotation=False)

Elasticly deform a batch. Requests larger batches upstream to avoid data loss due to rotation and jitter.

Parameters:

  • control_point_spacing (tuple of int) – Distance between control points for the elastic deformation, in voxels per dimension.

  • control_point_displacement_sigma (tuple of float) – Standard deviation of control point displacement distribution, in world coordinates.

  • rotation_interval (tuple of two floats) – Interval to randomly sample rotation angles from (0, 2PI).

  • subsample (int) – Instead of creating an elastic transformation on the full resolution, create one sub-sampled by the given factor, and linearly interpolate to obtain the full resolution transformation. This can significantly speed up this node, at the expense of having visible piecewise linear deformations for large factors. Usually, a factor of 4 can safely be used without noticeable changes. However, the default is 1 (i.e., no sub-sampling).

  • seed (int) – Set random state for reproducible results (tests only, do not use in production code!!)

setup()

To be implemented in subclasses.

Called during initialization of the DAG. Callees can assume that all upstream providers are set up already.

In setup, call provides() or updates() to announce the arrays and points provided or changed by this node.

prepare(request)

To be implemented in subclasses.

Prepare for a batch request. Should return a BatchRequest of needed dependencies. If None is returned, it will be assumed that all of request is needed.

process(batch, request)

To be implemented in subclasses.

Filter a batch, will be called after prepare(). Should return a Batch containing modified Arrays and Graphs. Keys in the returned batch will replace the associated data in the original batch. If None is returned it is assumed that the batch has been modified in place. request is the same as passed to prepare(), provided for convenience.

Parameters:

  • batch (Batch) – The batch received from upstream to be modified by this node.

  • request (BatchRequest) – The request this node received. The updated batch should meet this request.