API reference

class napari_correct_drift.ArrayAxesStandardizer(out_order, in_order)
This class standardizes the axes of numpy arrays by rearranging

the axes order.

Axes orders are given by strings, e. g. ‘tzcyx’

out_order

A string representing the desired output order of the array axes.

Type

str

in_order

A string representing the current order of the array axes.

Type

str

__call__(data)

Standardizes given data array

Takes an input numpy array and standardizes the axes according to the output order. It returns the standardized array.

Parameters

data (np.array) – A numpy array that needs to be standardized.

Returns

A standardized numpy array.

Return type

np.array

Raises

AssertionError – If the shape of the input array does not match the input order.

__init__(out_order, in_order)

Initializes the class and checks for any invalid inputs.

Parameters

  • out_order (str) – A string representing the desired output order of the array axes.

  • in_order (str) – A string representing the current order of the array axes.

Raises

AssertionError – If in_order contains any elements not in out_order or if out_order or in_order contains duplicates.

inv(data)

Inverse standardization

Applies the inverse transform to the input data. The input data is expected to be a numpy array with the shape specified by the ‘out_order’ argument passed to the constructor. The output is a numpy array with the shape specified by the ‘in_order’ argument passed to the constructor.

Parameters

data (np.array) – The input data to be transformed.

Returns

The transformed data.

Return type

np.array

Raises

  • AssertionError – If the input data does not have the same shape

  • as specified by the 'out_order' argument. –

class napari_correct_drift.CorrectDrift(data, dims)

Main drift correction class

is_multi_channel

True if the data has multiple channels.

Type

bool

is_3d

True if the data has a z dimension.

Type

bool

dims

The dimension order of the input data.

Type

str

data_arranger

An ArrayAxesStandardizer instance used to rearrange the input data.

Type

ArrayAxesStandardizer

data

The rearranged input data.

Type

np.array

T, C, Z, Y, X

The number of time points, channels, z-slices, rows, and columns in the input data.

Type

int

__init__(data, dims)

Init of main drift correction class

Parameters

  • data (np.array) – image data

  • dims (str) – corresponding axes

apply_drifts(offsets, extend_output=False, order=1, mode='constant')

Apply drifts to effectively create the corrected output image

Parameters

  • offsets (np.array) – the drift table

  • extend_output (bool, optional) – if True, the output array is extended to always contain the original frame. Defaults to False.

  • order (int, optional) – order of interpolation (1=linear). Defaults to 1.

  • mode (str, optional) – border treatment. Defaults to “constant”.

Returns

the corrected output image

Return type

np.array

estimate_drift(t0=0, channel=0, increment=1, upsample_factor=1, roi=None, normalization='phase', mode='relative', max_shifts=None, use_window=True)

Estimate drift entry point.

Calls _estimate_drift_relative or _estimate_drift_absolute depending on given mode

Parameters

  • t0 (int, optional) – key frame. Defaults to 0.

  • channel (int, optional) – key channel. Defaults to 0.

  • increment (int, optional) – frame increment. Defaults to 1.

  • upsample_factor (int, optional) – upsample factor. Defaults to 1.

  • roi (ROIRect, optional) – ROI. Defaults to None.

  • normalization (str, optional) – normalization. Defaults to “phase”.

  • mode (str, optional) – mode of drift correction. Defaults to “relative”.

  • max_shifts (Tuple[int, int, int], optional) – maximum allowed shifts. Defaults to None.

  • use_winsow (bool) – use Hann window to prevent spectral leakage. Defautls to True.

Raises

AttributeError – if mode is not supported

interpolate_drift(offsets)

Linear interpolation of missing values given as np.nan values

Parameters

offsets (np.array) – The drift table containing missing values (np.nan)

Raises

RuntimeError – When too few values for interpolation

Returns

offsets with linearly interpolated values

Return type

offsets (np.array)

class napari_correct_drift.CorrectDriftDock(napari_viewer)

Main user interface for Napari-Correct-Drift.

The main widget is structured in several groups:

  1. Input Axes: Select napari layer to process

  2. Correct Drift: Estimate, load or correct drift

  3. Key Frames: Set the key frames and mode for correction

  4. Parameters: Set Parameters of the drift correction

  5. Outliers: Handle outlier drifts

__init__(napari_viewer)
class napari_correct_drift.ROIRect(x_min, x_max, y_min, y_max, z_min, z_max, t0, c0)

Helper classes for 3D bounding-box, localized in channels and time

x_min, x_max, y_min, y_max, z_min, z_max

special boundaries of the ROI rectangle.

Type

int

t0

frame index

Type

int

c0

channel index

Type

int

__init__(x_min, x_max, y_min, y_max, z_min, z_max, t0, c0)
classmethod from_bbox(bbox, t0, c0)

Creates ROIRect from from bounding box list:

[z_min, z_max, y_min, y_max, x_min, x_max,]

Parameters

  • bbox (np.array) – bounding box

  • t0 (int) – key frame

  • c0 (int) – key channel

classmethod from_shape_poly(shape_poly, dims, z_min, z_max)

Creates ROIRect from napari shape polygon

Parameters

  • shape_poly (np.array) – napari polygon shape

  • dims (str) – dimensions string

  • z_min (int) – z min

  • z_max (int) – z max

class napari_correct_drift.TableWidget(ist=None, viewer=None)

A editable table widget containing the estimated drifts.

The underlying table data can be saved or copied to clipboard.

__init__(ist=None, viewer=None)
napari_correct_drift.sample_2d()

Creates 2D + time example data

napari_correct_drift.sample_3d()

Creates 3D + time example data

napari_correct_drift.sample_3d_ch()

Creates 2D + channel + time example data

napari_correct_drift.window_nd(shape, win_func=<function hann>)

N-dimensional windowing using 1d win_func. Default Hann window.

Note 1: shape should be at least 2d

Note 2: this function is cached.

Parameters

  • shape (Tuple[int, ...]) – Desired output shape

  • win_func (Callable, optional) – 1d Window function as in scipy.signal.windows. Defaults to windows.hann.

Returns

Array with applied window function.

Return type

np.array

Return type

array