gridcells.analysis.registration - Positional data registration.

Use the classes here to align (register) positional data of several recordings with the specified arena coordinates.

Classes

ArenaOriginRegistration([arena]) Register positional data to zero-coordinates of an arena.
OriginRegistrationResult(positions, offsets) A holder for registered data.
class gridcells.analysis.registration.ArenaOriginRegistration(arena=None)[source]

Bases: object

Register positional data to zero-coordinates of an arena.

The actual positional data recordings are prone to outliers. This registration engine ensures that the positional data from different recordings are “aligned” with respect to the arena coordinates. This is accomplished by optimising the positional offsets with respect to the number of outliers.

Todo

Deal with rotations.

Initialise with an arena against which to register the data.

Also use set_arena() to change the specific arena.

__init__(arena=None)[source]

Initialise with an arena against which to register the data.

Also use set_arena() to change the specific arena.

register(positions)[source]

Register the positional data against the current arena.

Parameters:

positions : Position2D

Positional data.

Returns:

res : OriginRegistrationResult

The result object, containing new positional data and the determined offsets.

set_arena(arena)[source]

Set the arena for registration.

All subsequent calls to register() will be performed on this arena.

class gridcells.analysis.registration.OriginRegistrationResult(positions, offsets)[source]

Bases: object

A holder for registered data.

Contains two attributes: positions and estimated offsets in the arena.