Extract

class celltk.extract.Extract(images=[], masks=[], channels=[], regions=[], lineages=[], time=None, condition='condition', position_id=0, min_trace_length=0, remove_parent=True, output='data_frame', save=True, force_rerun=True, skip_frames=(), _output_id=None, **kwargs)
Parameters

  • images (Collection[str], default: []) –

  • masks (Collection[str], default: []) –

  • channels (Collection[str], default: []) –

  • regions (Collection[str], default: []) –

  • lineages (Collection[ndarray], default: []) –

  • time (Optional[float], default: None) –

  • condition (str, default: 'condition') –

  • position_id (int, default: 0) –

  • min_trace_length (int, default: 0) –

  • remove_parent (bool, default: True) –

  • output (str, default: 'data_frame') –

  • save (bool, default: True) –

  • force_rerun (bool, default: True) –

  • skip_frames (Tuple[int], default: ()) –

  • _output_id (Optional[Tuple[str]], default: None) –

add_derived_metric(metric_name, keys, func='sum', inverse=False, propagate=False, frame_rng=None, *args, **kwargs)

Calculate additional metrics based on information already in array

Parameters

  • metric_name (str) – Key to save the metric under

  • keys (Collection[Tuple[str]]) – One or multiple keys to calculate with. Each key will be used to index an array, ConditionArray[key]. Each key should produce a 2D array when indexed.

  • func (str, default: 'sum') – Name of numpy function to apply, e.g. ‘sum’

  • inverse (bool, default: False) – If True, repeats the calculation as if the keys were passed in the opposite order and saves in the other keys.

  • propagate ((str, bool), default: False) – If True, propagates the results of the calculation to the other keys in the array.

  • frame_rng (Union[int, Tuple[int], None], default: None) – Frames to use in calculation. If int, takes that many frames from start of trace. If tuple, uses passed frames.

Return type

None

add_extra_metric(name, func=None)

Add custom metrics or metrics from regionprops to array. If function is none, value will just be nan.

Parameters

  • name (str) – key for the metric in the ConditionArray

  • func (Optional[Callable], default: None) – If str, name of the metric from skiamge.regionprops. if Callable, function that calculates the metric. Cannot be used if Operation must be saved as YAML before running.

Return type

None

add_filter(filter_name, metric, region=0, channel=0, frame_rng=None, *args, **kwargs)

Remove cells from array that do not match the filter.

Parameters

  • filter_name (str) – Options are ‘outside’, ‘inside’, ‘outside_percentile’, ‘inside_percentile’.

  • metric (str) – Name of metric to use. Can be any key in the array.

  • region (Union[str, int], default: 0) – Name of region to calculate the filter in.

  • channel (Union[str, int], default: 0) – Name of channel to calculate filter in.

  • frame_rng (Union[int, Tuple[int], None], default: None) – Frames to use in calculation. If int, takes that many frames from start of trace. If tuple, uses passed frames.

Return type

None

extract_data_from_image(images, masks, channels=[], regions=[], lineages=[], time=None, condition='default', position_id=None, min_trace_length=0, skip_frames=(), remove_parent=True, parent_track=0)

Extracts data from stacks of images and constructs a ConditionArray.

Parameters

  • images (Image) – Images to extract data from.

  • masks (Mask) – Masks to segment images with.

  • channels (Collection[str], default: []) – Names of channels corresponding to

  • regions (Collection[str], default: []) – Names of segmented regions corresponding, tracks and masks, in that order.

  • lineages (Collection[ndarray], default: []) – Lineage files corresponding to masks if provided.

  • time (Union[float, ndarray, None], default: None) – If int or float, designates time between frames. If array, marks the frame time points.

  • condition (str, default: 'default') – Name of the condition

  • position_id (Optional[int], default: None) – Unique identifier if multiple ConditionArrays will share the same condition

  • min_trace_length (int, default: 0) – All cells with shorter traces will be deleted from the final array.

  • skip_frames (Tuple[int], default: ()) – Use to specify frames to be skipped. If provided to Pipeline, does not need to be provided again, but must match.

  • remove_parent (bool, default: True) – If true, parents of cells are not kept in the final ConditionArray.

  • parent_track (int, default: 0) – If multiple tracks are provided, designates the one to use for lineage tracking

Returns

ConditionArray with data from the images.

Return type

ConditionArray

The following axes dimensions are used: ax 0 - cell locations (nuc, cyto, population, etc.) ax 1 - channels (TRITC, FITC, etc.) ax 2 - metrics (median_int, etc.) ax 3 - cells ax 4 - frames

set_metric_list(metrics)

Sets the list of metrics to get. For a possible list, see skimage.regionprops or Extract._possible_metrics.

Parameters

metrics (Collection[str]) – List of metrics to measure from images

Return type

None

Returns

None

Note

  • CellTK can only use the scalar metrics in regionprops.