piscis.core =========== .. py:module:: piscis.core Classes ------- .. autoapisummary:: piscis.core._Piscis piscis.core.Piscis piscis.core.PiscisLegacy Functions --------- .. autoapisummary:: piscis.core.adjust_parameters Module Contents --------------- .. py:class:: _Piscis(model_name: str, batch_size: int, input_size: Optional[Tuple[int, int]], device: Optional[Union[str, torch.device]], pooling: str) Base class for running the Piscis algorithm. :Attributes: **model_name** : str Model name. **batch_size** : int Batch size for the CNN. **adjustment** : str Adjustment type applied to images during preprocessing. **input_size** : Tuple[int, int] Input size for the CNN. **dilation_iterations** : int Number of iterations used to dilate ground truth labels during training. **channels** : int Number of channels in the input images. **device** : Optional[Union[str, torch.device]] Device to run the model on. **pooling** : str Pooling type applied to labels. **model** : SpotsModel Model. .. !! processed by numpydoc !! .. py:attribute:: model_name .. py:attribute:: batch_size .. py:attribute:: device .. py:attribute:: pooling .. py:attribute:: model .. py:method:: _predict(x: Union[numpy.ndarray, dask.array.Array], stack: bool, scale: float, threshold: float, min_distance: int, intermediates: bool) -> Union[Tuple[numpy.ndarray, xarray.DataArray], numpy.ndarray] Predict spots in an image or stack of images. :Parameters: **x** : np.ndarray or da.Array Image or stack of images. **stack** : bool, optional Whether `x` is a stack of images. **scale** : float, optional Scale factor for rescaling `x`. **threshold** : float, optional Spot detection threshold. **min_distance** : int, optional Minimum distance between spots. **intermediates** : bool, optional Whether to return intermediate feature maps. :Returns: **coords** : np.ndarray Predicted spot coordinates. **y** : np.ndarray, optional Intermediate feature maps. Only returned if `intermediates` is True. .. !! processed by numpydoc !! .. py:method:: _predict_stack(tiles: deeptile.core.data.Tiled, stack_axis_len: int, scales: numpy.ndarray, threshold: float, min_distance: int, intermediates: bool) -> Union[Tuple[deeptile.Output, numpy.ndarray], deeptile.Output] Predict spots in a stack of images. :Parameters: **tiles** : Tiled Tiles of images. **stack_axis_len** : int Length of the stack axis. **scales** : np.ndarray Scales for rescaling tiles. **threshold** : float Spot detection threshold. Can be interpreted as the minimum number of fully confident pixels necessary to identify a spot. **min_distance** : int Minimum distance between spots. **intermediates** : bool Whether to return intermediate feature maps. :Returns: **coords** : Output Predicted spot coordinates. **y** : np.ndarray, optional Intermediate feature maps. Only returned if `intermediates` is True. .. !! processed by numpydoc !! .. py:method:: _predict_plane(tiles: deeptile.core.data.Tiled, scales: numpy.ndarray, threshold: float, min_distance: int, intermediates: bool) -> Union[Tuple[deeptile.Output, numpy.ndarray], deeptile.Output] Predict spots in a single plane. :Parameters: **tiles** : Tiled Tiles of images. **scales** : np.ndarray Scales for rescaling tiles. **threshold** : float Spot detection threshold. Can be interpreted as the minimum number of fully confident pixels necessary to identify a spot. **min_distance** : int Minimum distance between spots. **intermediates** : bool Whether to return intermediate feature maps. :Returns: **coords** : Output Predicted spot coordinates. **y** : np.ndarray, optional Intermediate feature maps. Only returned if `intermediates` is True. .. !! processed by numpydoc !! .. py:method:: _preprocess(x: Union[numpy.ndarray, dask.array.Array], stack: bool) -> Tuple[dask.array.Array, bool] Preprocess the input. :Parameters: **x** : np.ndarray or dask.array.Array Image or stack of images. **stack** : bool Whether `x` is a stack of images. :Returns: **x** : Union[np.ndarray, da.Array] Preprocessed image or stack of images. **batch_axis** : bool Whether `x` has a batch axis. :Raises: ValueError If `x` does not have the correct dimensions. ValueError If `x` does not have the correct number of channels. .. !! processed by numpydoc !! .. py:method:: _process(tiles: deeptile.core.data.Tiled) -> Tuple[numpy.ndarray, numpy.ndarray] Process tiles. :Parameters: **tiles** : Tiled Tiles of images. :Returns: **labels** : np.ndarray Predicted labels. **deltas** : np.ndarray Predicted displacement vectors. .. !! processed by numpydoc !! .. py:method:: _postprocess_stack(labels: numpy.ndarray, deltas: numpy.ndarray, scales: numpy.ndarray, threshold: float, min_distance: int) -> deeptile.Output :staticmethod: Postprocess tiles. :Parameters: **labels** : np.ndarray Predicted labels. **deltas** : np.ndarray Predicted displacement vectors. **scales** : np.ndarray Scales for rescaling tiles. **threshold** : float Spot detection threshold. Can be interpreted as the minimum number of fully confident pixels necessary to identify a spot. **min_distance** : int Minimum distance between spots. :Returns: **coords** : Output Predicted spot coordinates. .. !! processed by numpydoc !! .. py:method:: _apply(x: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor] Apply SpotsModel to a batch images. x : torch.Tensor Batch of images. :Returns: **labels** : np.ndarray Predicted binary labels. **deltas** : np.ndarray Predicted displacements vectors. .. !! processed by numpydoc !! .. py:class:: Piscis(model_name: str = '20251212', batch_size: int = 1, input_size: Optional[Tuple[int, int]] = None, device: Optional[Union[str, torch.device]] = None) Bases: :py:obj:`_Piscis` Class for running the Piscis algorithm. :Attributes: **model_name** : str Model name. **batch_size** : int Batch size for the CNN. **adjustment** : str Adjustment type applied to images during preprocessing. **input_size** : Tuple[int, int] Input size for the CNN. **dilation_iterations** : int Number of iterations used to dilate ground truth labels during training. **channels** : int Number of channels in the input images. **device** : Optional[Union[str, torch.device]] Device to run the model on. **pooling** : str Pooling type applied to labels. **model** : SpotsModel Model. .. !! processed by numpydoc !! .. py:method:: predict(x: Union[numpy.ndarray, dask.array.Array], stack: bool = False, scale: float = 1.0, threshold: float = 0.5, min_distance: int = 1, intermediates: bool = False) -> Union[Tuple[numpy.ndarray, xarray.DataArray], numpy.ndarray] Predict spots in an image or stack of images. :Parameters: **x** : np.ndarray or da.Array Image or stack of images. **stack** : bool, optional Whether `x` is a stack of images. Default is False. **scale** : float, optional Scale factor for rescaling `x`. Default is 1. **threshold** : float, optional Spot detection threshold. Default is 0.5. **min_distance** : int, optional Minimum distance between spots. Default is 1. **intermediates** : bool, optional Whether to return intermediate feature maps. Default is False. :Returns: **coords** : np.ndarray Predicted spot coordinates. **y** : np.ndarray, optional Intermediate feature maps. Only returned if `intermediates` is True. .. !! processed by numpydoc !! .. py:class:: PiscisLegacy(model_name: str = '20230905', batch_size: int = 1, input_size: Optional[Tuple[int, int]] = None, device: Optional[Union[str, torch.device]] = None) Bases: :py:obj:`_Piscis` Class for running the PyTorch port of the legacy Piscis algorithm originally implemented in JAX. :Attributes: **model_name** : str Model name. **batch_size** : int Batch size for the CNN. **adjustment** : str Adjustment type applied to images during preprocessing. **input_size** : Tuple[int, int] Input size for the CNN. **dilation_iterations** : int Number of iterations used to dilate ground truth labels during training. **channels** : int Number of channels in the input images. **device** : Optional[Union[str, torch.device]] Device to run the model on. **pooling** : str Pooling type applied to labels. **model** : SpotsModel Model. .. !! processed by numpydoc !! .. py:method:: predict(x: Union[numpy.ndarray, dask.array.Array], stack: bool = False, scale: float = 1.0, threshold: float = 1.0, min_distance: int = 1, intermediates: bool = False) -> Union[Tuple[numpy.ndarray, xarray.DataArray], numpy.ndarray] Predict spots in an image or stack of images. :Parameters: **x** : np.ndarray or da.Array Image or stack of images. **stack** : bool, optional Whether `x` is a stack of images. Default is False. **scale** : float, optional Scale factor for rescaling `x`. Default is 1. **threshold** : float, optional Spot detection threshold. Can be interpreted as the minimum number of fully confident pixels necessary to identify a spot. Default is 1.0. **min_distance** : int, optional Minimum distance between spots. Default is 1. **intermediates** : bool, optional Whether to return intermediate feature maps. Default is False. :Returns: **coords** : np.ndarray Predicted spot coordinates. **y** : np.ndarray, optional Intermediate feature maps. Only returned if `intermediates` is True. .. !! processed by numpydoc !! .. py:function:: adjust_parameters(y: xarray.DataArray, threshold: float = 0.5, min_distance: int = 1) -> numpy.ndarray Adjust tunable parameters for a given set of intermediate feature maps. :Parameters: **y** : xr.DataArray Intermediate feature maps. **threshold: float** Spot detection threshold. Default is 0.5. **min_distance** : int, optional Minimum distance between spots. Default is 1. :Returns: **coords** : np.ndarray Predicted spot coordinates. .. !! processed by numpydoc !!