junifer 0.0.6.dev154__py3-none-any.whl → 0.0.6.dev194__py3-none-any.whl

junifer/__init__.pyi

@@ -12,6 +12,7 @@ __all__ = [
     "utils",
     "external",
     "onthefly",
+    "testing",
     "__version__",
 ]
 
@@ -28,6 +29,7 @@ from . import (
     storage,
     utils,
     external,
+    testing,
     onthefly,
 )
 from ._version import __version__

junifer/_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.0.6.dev154'
-__version_tuple__ = version_tuple = (0, 0, 6, 'dev154')
+__version__ = version = '0.0.6.dev194'
+__version_tuple__ = version_tuple = (0, 0, 6, 'dev194')

junifer/api/decorators.py

@@ -7,7 +7,7 @@
 
 from typing import Type
 
-from ..pipeline.registry import register
+from ..pipeline import PipelineComponentRegistry
 
 
 __all__ = [
@@ -39,9 +39,8 @@ def register_datagrabber(klass: Type) -> Type:
     It should only be used as a decorator.
 
     """
-    register(
+    PipelineComponentRegistry().register(
         step="datagrabber",
-        name=klass.__name__,
         klass=klass,
     )
     return klass
@@ -67,9 +66,8 @@ def register_datareader(klass: Type) -> Type:
     It should only be used as a decorator.
 
     """
-    register(
+    PipelineComponentRegistry().register(
         step="datareader",
-        name=klass.__name__,
         klass=klass,
     )
     return klass
@@ -91,9 +89,8 @@ def register_preprocessor(klass: Type) -> Type:
         The unmodified input class.
 
     """
-    register(
+    PipelineComponentRegistry().register(
         step="preprocessing",
-        name=klass.__name__,
         klass=klass,
     )
     return klass
@@ -115,9 +112,8 @@ def register_marker(klass: Type) -> Type:
         The unmodified input class.
 
     """
-    register(
+    PipelineComponentRegistry().register(
         step="marker",
-        name=klass.__name__,
         klass=klass,
     )
     return klass
@@ -139,9 +135,8 @@ def register_storage(klass: Type) -> Type:
         The unmodified input class.
 
     """
-    register(
+    PipelineComponentRegistry().register(
         step="storage",
-        name=klass.__name__,
         klass=klass,
     )
     return klass

junifer/api/functions.py

@@ -7,15 +7,17 @@
 
 import os
 import shutil
-import typing
 from pathlib import Path
 from typing import Dict, List, Optional, Tuple, Union
 
 from ..api.queue_context import GnuParallelLocalAdapter, HTCondorAdapter
 from ..datagrabber.base import BaseDataGrabber
 from ..markers.base import BaseMarker
-from ..pipeline import MarkerCollection, WorkDirManager
-from ..pipeline.registry import build
+from ..pipeline import (
+    MarkerCollection,
+    PipelineComponentRegistry,
+    WorkDirManager,
+)
 from ..preprocess.base import BasePreprocessor
 from ..storage.base import BaseFeatureStorage
 from ..utils import logger, raise_error, yaml
@@ -38,42 +40,78 @@ def _get_datagrabber(datagrabber_config: Dict) -> BaseDataGrabber:
         The DataGrabber.
 
     """
-    datagrabber_params = datagrabber_config.copy()
-    datagrabber_kind = datagrabber_params.pop("kind")
-    datagrabber = build(
+    return PipelineComponentRegistry().build_component_instance(
         step="datagrabber",
-        name=datagrabber_kind,
+        name=datagrabber_config.pop("kind"),
         baseclass=BaseDataGrabber,
-        init_params=datagrabber_params,
+        init_params=datagrabber_config,
     )
-    datagrabber = typing.cast(BaseDataGrabber, datagrabber)
-    return datagrabber
 
 
 def _get_preprocessor(preprocessing_config: Dict) -> BasePreprocessor:
-    """Get preprocessor.
+    """Get Preprocessor.
 
     Parameters
     ----------
     preprocessing_config : dict
-        The config to get the preprocessor using.
+        The config to get the Preprocessor using.
 
     Returns
     -------
     dict
-        The preprocessor.
+        The Preprocessor.
 
     """
-    preprocessor_params = preprocessing_config.copy()
-    preprocessor_kind = preprocessor_params.pop("kind")
-    preprocessor = build(
+    return PipelineComponentRegistry().build_component_instance(
         step="preprocessing",
-        name=preprocessor_kind,
+        name=preprocessing_config.pop("kind"),
         baseclass=BasePreprocessor,
-        init_params=preprocessor_params,
+        init_params=preprocessing_config,
+    )
+
+
+def _get_marker(marker_config: Dict) -> BaseMarker:
+    """Get Marker.
+
+    Parameters
+    ----------
+    marker_config : dict
+        The config to get the Marker using.
+
+    Returns
+    -------
+    object
+        The Marker.
+
+    """
+    return PipelineComponentRegistry().build_component_instance(
+        step="marker",
+        name=marker_config.pop("kind"),
+        baseclass=BaseMarker,
+        init_params=marker_config,
+    )
+
+
+def _get_storage(storage_config: Dict) -> BaseFeatureStorage:
+    """Get Storage.
+
+    Parameters
+    ----------
+    storage_config : dict
+        The config to get the Storage using.
+
+    Returns
+    -------
+    dict
+        The Storage.
+
+    """
+    return PipelineComponentRegistry().build_component_instance(
+        step="storage",
+        name=storage_config.pop("kind"),
+        baseclass=BaseFeatureStorage,
+        init_params=storage_config,
     )
-    preprocessor = typing.cast(BasePreprocessor, preprocessor)
-    return preprocessor
 
 
 def run(
@@ -123,41 +161,22 @@ def run(
         elements = [elements]
 
     # Get datagrabber to use
-    datagrabber_object = _get_datagrabber(datagrabber)
+    datagrabber_object = _get_datagrabber(datagrabber.copy())
 
-    # Copy to avoid changing the original dict
-    _markers = [x.copy() for x in markers]
-    built_markers = []
-    for t_marker in _markers:
-        kind = t_marker.pop("kind")
-        t_m = build(
-            step="marker",
-            name=kind,
-            baseclass=BaseMarker,
-            init_params=t_marker,
-        )
-        built_markers.append(t_m)
+    # Get markers to use
+    built_markers = [_get_marker(marker) for marker in markers.copy()]
 
     # Get storage engine to use
-    storage_params = storage.copy()
-    storage_kind = storage_params.pop("kind")
-    if "single_output" not in storage_params:
-        storage_params["single_output"] = False
-    storage_object = build(
-        step="storage",
-        name=storage_kind,
-        baseclass=BaseFeatureStorage,
-        init_params=storage_params,
-    )
-    storage_object = typing.cast(BaseFeatureStorage, storage_object)
+    if "single_output" not in storage:
+        storage["single_output"] = False
+    storage_object = _get_storage(storage.copy())
 
     # Get preprocessor to use (if provided)
     if preprocessors is not None:
-        _preprocessors = [x.copy() for x in preprocessors]
-        built_preprocessors = []
-        for preprocessor in _preprocessors:
-            preprocessor_object = _get_preprocessor(preprocessor)
-            built_preprocessors.append(preprocessor_object)
+        built_preprocessors = [
+            _get_preprocessor(preprocessor)
+            for preprocessor in preprocessors.copy()
+        ]
     else:
         built_preprocessors = None
 
@@ -167,6 +186,7 @@ def run(
         preprocessors=built_preprocessors,
         storage=storage_object,
     )
+    # Validate the marker collection for the datagrabber
     mc.validate(datagrabber_object)
 
     # Fit elements
@@ -192,19 +212,11 @@ def collect(storage: Dict) -> None:
         constructor.
 
     """
-    storage_params = storage.copy()
-    storage_kind = storage_params.pop("kind")
-    logger.info(f"Collecting data using {storage_kind}")
-    logger.debug(f"\tStorage params: {storage_params}")
-    if "single_output" not in storage_params:
-        storage_params["single_output"] = False
-    storage_object = build(
-        step="storage",
-        name=storage_kind,
-        baseclass=BaseFeatureStorage,
-        init_params=storage_params,
-    )
-    storage_object = typing.cast(BaseFeatureStorage, storage_object)
+    logger.info(f"Collecting data using {storage['kind']}")
+    logger.debug(f"\tStorage params: {storage}")
+    if "single_output" not in storage:
+        storage["single_output"] = False
+    storage_object = _get_storage(storage.copy())
     logger.debug("Running storage.collect()")
     storage_object.collect()
     logger.info("Collect done")

junifer/api/tests/test_functions.py

@@ -15,7 +15,7 @@ from ruamel.yaml import YAML
 import junifer.testing.registry  # noqa: F401
 from junifer.api import collect, list_elements, queue, reset, run
 from junifer.datagrabber.base import BaseDataGrabber
-from junifer.pipeline.registry import build
+from junifer.pipeline import PipelineComponentRegistry
 
 
 # Configure YAML class
@@ -240,7 +240,7 @@ def test_run_and_collect(
         storage=storage,
     )
     # Get datagrabber
-    dg = build(
+    dg = PipelineComponentRegistry().build_component_instance(
         step="datagrabber", name=datagrabber["kind"], baseclass=BaseDataGrabber
     )
     elements = dg.get_elements()  # type: ignore

junifer/data/__init__.pyi

@@ -1,41 +1,27 @@
 __all__ = [
-    "list_coordinates",
-    "load_coordinates",
-    "register_coordinates",
-    "get_coordinates",
-    "list_parcellations",
-    "load_parcellation",
-    "register_parcellation",
-    "merge_parcellations",
-    "get_parcellation",
-    "list_masks",
-    "load_mask",
-    "register_mask",
-    "get_mask",
+    "CoordinatesRegistry",
+    "ParcellationRegistry",
+    "MaskRegistry",
+    "get_data",
+    "list_data",
+    "load_data",
+    "register_data",
+    "deregister_data",
     "get_template",
     "get_xfm",
     "utils",
 ]
 
-from .coordinates import (
-    list_coordinates,
-    load_coordinates,
-    register_coordinates,
-    get_coordinates,
-)
-from .parcellations import (
-    list_parcellations,
-    load_parcellation,
-    register_parcellation,
-    merge_parcellations,
-    get_parcellation,
-)
+from .coordinates import CoordinatesRegistry
+from .parcellations import ParcellationRegistry
+from .masks import MaskRegistry
 
-from .masks import (
-    list_masks,
-    load_mask,
-    register_mask,
-    get_mask,
+from ._dispatch import (
+    get_data,
+    list_data,
+    load_data,
+    register_data,
+    deregister_data,
 )
 
 from .template_spaces import get_template, get_xfm

junifer/data/_dispatch.py

@@ -0,0 +1,251 @@
+"""Provide dispatch functions for pipeline data registries."""
+
+# Authors: Synchon Mandal <s.mandal@fz-juelich.de>
+# License: AGPL
+
+from pathlib import Path
+from typing import (
+    TYPE_CHECKING,
+    Any,
+    Callable,
+    Dict,
+    List,
+    Optional,
+    Tuple,
+    Union,
+)
+
+from numpy.typing import ArrayLike
+
+from ..utils import raise_error
+from .coordinates import CoordinatesRegistry
+from .masks import MaskRegistry
+from .parcellations import ParcellationRegistry
+
+
+if TYPE_CHECKING:
+    from nibabel.nifti1 import Nifti1Image
+
+
+__all__ = [
+    "get_data",
+    "list_data",
+    "load_data",
+    "register_data",
+    "deregister_data",
+]
+
+
+def get_data(
+    kind: str,
+    names: Union[
+        str,  # coordinates, parcellation, mask
+        List[str],  # parcellation, mask
+        Dict,  # mask
+        List[Dict],  # mask
+    ],
+    target_data: Dict[str, Any],
+    extra_input: Optional[Dict[str, Any]] = None,
+) -> Union[
+    Tuple[ArrayLike, List[str]],  # coordinates
+    Tuple["Nifti1Image", List[str]],  # parcellation
+    "Nifti1Image",  # mask
+]:
+    """Get tailored ``kind`` for ``target_data``.
+
+    Parameters
+    ----------
+    kind : {"coordinates", "parcellation", "mask"}
+        Kind of data to fetch and apply.
+    names : str or dict or list of str / dict
+        The registered name(s) of the data.
+    target_data : dict
+        The corresponding item of the data object to which the data
+        will be applied.
+    extra_input : dict, optional
+        The other fields in the data object. Useful for accessing other
+        data types that need to be used in the computation of data
+        (default None).
+
+    Returns
+    -------
+    tuple of numpy.ndarray, list of str; \
+    tuple of nibabel.nifti1.Nifti1Image, list of str; \
+    nibabel.nifti1.Nifti1Image
+
+    Raises
+    ------
+    ValueError
+        If ``kind`` is invalid value.
+
+    """
+
+    if kind == "coordinates":
+        return CoordinatesRegistry().get(
+            coords=names,
+            target_data=target_data,
+            extra_input=extra_input,
+        )
+    elif kind == "parcellation":
+        return ParcellationRegistry().get(
+            parcellations=names,
+            target_data=target_data,
+            extra_input=extra_input,
+        )
+    elif kind == "mask":
+        return MaskRegistry().get(
+            masks=names,
+            target_data=target_data,
+            extra_input=extra_input,
+        )
+    else:
+        raise_error(f"Unknown data kind: {kind}")
+
+
+def list_data(kind: str) -> List[str]:
+    """List available data for ``kind``.
+
+    Parameters
+    ----------
+    kind : {"coordinates", "parcellation", "mask"}
+        Kind of data registry to list.
+
+    Returns
+    -------
+    list of str
+        Available data for the registry.
+
+    Raises
+    ------
+    ValueError
+        If ``kind`` is invalid value.
+
+    """
+
+    if kind == "coordinates":
+        return CoordinatesRegistry().list
+    elif kind == "parcellation":
+        return ParcellationRegistry().list
+    elif kind == "mask":
+        return MaskRegistry().list
+    else:
+        raise_error(f"Unknown data kind: {kind}")
+
+
+def load_data(
+    kind: str,
+    name: str,
+    **kwargs,
+) -> Union[
+    Tuple[ArrayLike, List[str], str],  # coordinates
+    Tuple[Optional["Nifti1Image"], List[str], Path, str],  # parcellation
+    Tuple[
+        Optional[Union["Nifti1Image", Callable]], Optional[Path], str
+    ],  # mask
+]:
+    """Load ``kind`` named ``name``.
+
+    Parameters
+    ----------
+    kind : {"coordinates", "parcellation", "mask"}
+        Kind of data to load.
+    name : str
+        The registered name of the data.
+    **kwargs
+        Keyword arguments are passed to respective registry class method.
+
+    Returns
+    -------
+    tuple of numpy.ndarray, list of str, str; \
+    tuple of nibabel.nifti1.Nifti1Image or None, \
+    list of str, pathlib.Path, str; \
+    tuple of nibabel.nifti1.Nifti1Image or callable or None, \
+    pathlib.Path or None, str
+
+    Raises
+    ------
+    ValueError
+        If ``kind`` is invalid value.
+
+    """
+
+    if kind == "coordinates":
+        return CoordinatesRegistry().load(name=name)
+    elif kind == "parcellation":
+        return ParcellationRegistry().load(name=name, **kwargs)
+    elif kind == "mask":
+        return MaskRegistry().load(name=name, **kwargs)
+    else:
+        raise_error(f"Unknown data kind: {kind}")
+
+
+def register_data(
+    kind: str,
+    name: str,
+    space: str,
+    overwrite: bool = False,
+    **kwargs,
+) -> None:
+    """Register ``name`` under ``kind``.
+
+    Parameters
+    ----------
+    kind : {"coordinates", "parcellation", "mask"}
+        Kind of data to register.
+    name : str
+        The name to register.
+    space : str
+        The template space of the data.
+    overwrite : bool, optional
+        If True, overwrite an existing data with the same name.
+    **kwargs
+        Keyword arguments are passed to respective registry class method.
+
+    Raises
+    ------
+    ValueError
+        If ``kind`` is invalid value.
+
+    """
+
+    if kind == "coordinates":
+        return CoordinatesRegistry().register(
+            name=name, space=space, overwrite=overwrite, **kwargs
+        )
+    elif kind == "parcellation":
+        return ParcellationRegistry().register(
+            name=name, space=space, overwrite=overwrite, **kwargs
+        )
+    elif kind == "mask":
+        return MaskRegistry().load(
+            name=name, space=space, overwrite=overwrite, **kwargs
+        )
+    else:
+        raise_error(f"Unknown data kind: {kind}")
+
+
+def deregister_data(kind: str, name: str) -> None:
+    """De-register ``name`` from ``kind``.
+
+    Parameters
+    ----------
+    kind : {"coordinates", "parcellation", "mask"}
+        Kind of data to register.
+    name : str
+        The name to de-register.
+
+    Raises
+    ------
+    ValueError
+        If ``kind`` is invalid value.
+
+    """
+
+    if kind == "coordinates":
+        return CoordinatesRegistry().deregister(name=name)
+    elif kind == "parcellation":
+        return ParcellationRegistry().register(name=name)
+    elif kind == "mask":
+        return MaskRegistry().load(name=name)
+    else:
+        raise_error(f"Unknown data kind: {kind}")

junifer/data/coordinates/__init__.py

@@ -0,0 +1,9 @@
+"""Coordinates."""
+
+# Authors: Synchon Mandal <s.mandal@fz-juelich.de>
+# License: AGPL
+
+import lazy_loader as lazy
+
+
+__getattr__, __dir__, __all__ = lazy.attach_stub(__name__, __file__)

junifer/data/coordinates/__init__.pyi

@@ -0,0 +1,5 @@
+__all__ = [
+    "CoordinatesRegistry",
+]
+
+from ._coordinates import CoordinatesRegistry