Cellfinder workflow

MANIFEST.in

@@ -3,6 +3,7 @@ include README.md
 exclude .pre-commit-config.yaml
 
 recursive-include brainglobe_workflows *.py
+recursive-exclude brainglobe_workflows/cellfinder/notebooks *.py
 
 recursive-exclude * __pycache__
 recursive-exclude * *.py[co]

brainglobe_workflows/__init__.py

@@ -1,7 +1,7 @@
 from importlib.metadata import PackageNotFoundError, version
 
 try:
-    __version__ = version("brainglobe-scripts")
+    __version__ = version("brainglobe-workflows")
 except PackageNotFoundError:
     # package is not installed
     pass

brainglobe_workflows/cellfinder/cellfinder_main.py

@@ -0,0 +1,303 @@
+import datetime
+import json
+import logging
+import os
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Optional, Tuple, Union  # Any, Dict, Tuple,
+
+import pooch
+from brainglobe_utils.IO.cells import save_cells
+from cellfinder_core.main import main as cellfinder_run
+from cellfinder_core.tools.IO import read_with_dask
+from cellfinder_core.train.train_yml import depth_type
+
+Pathlike = Union[str, os.PathLike]  # Union[str, bytes, os.PathLike, Path]
+
+# Default config
+DATA_URL = "https://gin.g-node.org/BrainGlobe/test-data/raw/master/cellfinder/cellfinder-test-data.zip"
+DATA_HASH = "b0ef53b1530e4fa3128fcc0a752d0751909eab129d701f384fc0ea5f138c5914"
+CELLFINDER_CACHE_DIR = Path.home() / ".cellfinder_benchmarks"
+
+DEFAULT_CONFIG_DICT = {
+    "install_path": CELLFINDER_CACHE_DIR,
+    "data_url": DATA_URL,
+    "data_hash": DATA_HASH,
+    "local_path": CELLFINDER_CACHE_DIR / "cellfinder_test_data",
+    "signal_parent_dir": str(
+        CELLFINDER_CACHE_DIR / "cellfinder_test_data" / "signal"
+    ),
+    "background_parent_dir": str(
+        CELLFINDER_CACHE_DIR / "cellfinder_test_data" / "background"
+    ),
+    "output_path": CELLFINDER_CACHE_DIR / "cellfinder_output",
+    "detected_cells_filepath": (
+        CELLFINDER_CACHE_DIR / "cellfinder_output" / "detected_cells.xml"
+    ),
+    "voxel_sizes": [5, 2, 2],  # microns
+    "start_plane": 0,
+    "end_plane": -1,
+    "trained_model": None,  # if None, it will use a default model
+    "model_weights": None,
+    "model": "resnet50_tv",
+    "batch_size": 32,
+    "n_free_cpus": 2,
+    "network_voxel_sizes": [5, 1, 1],
+    "soma_diameter": 16,
+    "ball_xy_size": 6,
+    "ball_z_size": 15,
+    "ball_overlap_fraction": 0.6,
+    "log_sigma_size": 0.2,
+    "n_sds_above_mean_thresh": 10,
+    "soma_spread_factor": 1.4,
+    "max_cluster_size": 100000,
+    "cube_width": 50,
+    "cube_height": 50,
+    "cube_depth": 20,
+    "network_depth": "50",
+}
+
+
+@dataclass
+class CellfinderConfig:
+    """
+    Define input and output data locations, and parameters for
+    preprocessing steps.
+    """
+
+    # cellfinder benchmarks cache directory
+    install_path: Pathlike
+
+    # origin of data to download (if req)
+    data_url: Optional[str]  # either str or None
+    data_hash: Optional[str]
+
+    # cached subdirectory to save data to
+    local_path: Pathlike
+    signal_parent_dir: str
+    background_parent_dir: str
+    output_path: Pathlike
+    detected_cells_filepath: Pathlike
+
+    # preprocessing parameters
+    voxel_sizes: Tuple[float, float, float]
+    start_plane: int
+    end_plane: int
+    trained_model: Optional[
+        os.PathLike
+    ]  # if None, it will use a default model
+    model_weights: Optional[os.PathLike]
+    model: str
+    batch_size: int
+    n_free_cpus: int
+    network_voxel_sizes: Tuple[int, int, int]
+    soma_diameter: int
+    ball_xy_size: int
+    ball_z_size: int
+    ball_overlap_fraction: float
+    log_sigma_size: float
+    n_sds_above_mean_thresh: int
+    soma_spread_factor: float
+    max_cluster_size: int
+    cube_width: int
+    cube_height: int
+    cube_depth: int
+    network_depth: depth_type
+
+    list_signal_files: Optional[list] = None
+    list_background_files: Optional[list] = None
+
+
+def example_cellfinder_script():
+    cfg = setup_workflow()  # (this won't be timed)
+    run_workflow_from_cellfinder_run(cfg)  # (this will be timed)
+
+
+def run_workflow_from_cellfinder_run(cfg):
+    """
+    Run workflow based on the cellfinder_core.main.main()
+    function.
+
+    The steps are:
+    1. Read the input signal and background data as two separate
+       Dask arrays.
+    2. Run the main cellfinder pipeline on the input Dask arrays,
+       with the parameters defined in the input configuration (cfg).
+    3. Save the detected cells as an xml file to the location specified in
+       the input configuration (cfg).
+
+    We plan to time each of the steps in the workflow individually,
+    as well as the full workflow.
+
+    Parameters
+    ----------
+    cfg : CellfinderConfig
+        a class with the required setup methods and parameters for
+        the cellfinder workflow
+    """
+    # Read input data as Dask arrays
+    signal_array = read_with_dask(cfg.signal_parent_dir)
+    background_array = read_with_dask(cfg.background_parent_dir)
+
+    # Run main analysis using `cellfinder_run`
+    detected_cells = cellfinder_run(
+        signal_array, background_array, cfg.voxel_sizes
+    )
+
+    # Save results to xml file
+    save_cells(detected_cells, cfg.detected_cells_filepath)
+
+
+def setup_workflow():
+    """Prepare configuration to run workflow
+
+    This includes
+    - instantiating the config dictionary,
+    - checking if the input data exists locally, and fetching from
+      GIN repository otherwise,
+    - creating the directory for the output of the workflow if it doesn't exist
+
+    To instantiate the config dictionary, we first check if an environment
+    variable "CELLFINDER_CONFIG_PATH" pointing to a config json file exists.
+    If not, the default config (DEFAULT_CONFIG_DICT) is used.
+    Returns
+    -------
+    config : CellfinderConfig
+        a class with the required setup methods and parameters for
+        the cellfinder workflow
+    """
+
+    # Define config
+    if "CELLFINDER_CONFIG_PATH" in os.environ.keys():
+        input_config_path = Path(os.environ["CELLFINDER_CONFIG_PATH"])
+        assert input_config_path.exists()
+
+        # read into dict (assuming config is a json dict?)
+        # TODO:add error handling here?
+        with open(input_config_path) as cfg:
+            config_dict = json.load(cfg)
+
+        config = CellfinderConfig(**config_dict)
+
+        logging.info(
+            "Configuration retrieved from "
+            f'{os.environ["CELLFINDER_CONFIG_PATH"]}'
+        )
+
+    else:
+        config = CellfinderConfig(**DEFAULT_CONFIG_DICT)
+        logging.info("Using default configuration")
+
+    # Retrieve and add lists of input data to config if neither are defined
+    if not (config.list_signal_files and config.list_signal_files):
+        config = retrieve_input_data(config)
+
+    # Create output directory if it doesn't exist, timestamped
+    timestamp = datetime.datetime.now()
+    timestamp_formatted = timestamp.strftime("%Y%m%d_%H%M%S")
+    (Path(str(config.output_path) + "_" + timestamp_formatted)).mkdir(
+        parents=True, exist_ok=True
+    )
+
+    return config
+
+
+def retrieve_input_data(config):
+    """
+    Adds the lists of input data files (signal and background) to the config.
+
+    It first checks if the input data exists locally.
+    - If both directories (signal and background) exist, the lists of signal
+      and background files are added to the relevant config attributes
+    - If exactly one of the input data directories is missing, an error
+      message is logged.
+    - If neither of them exist, the data is retrieved from the provided GIN
+      repository. If no URL or hash to GIN is provided, an error is shown.
+
+    Parameters
+    ----------
+    config : CellfinderConfig
+        a dataclass whose attributes are the parameters
+        for running cellfinder.
+
+    Returns
+    -------
+    config : CellfinderConfig
+        a dataclass whose attributes are the parameters
+        for running cellfinder.
+    """
+    # Check if input data (signal and background) exist locally.
+    # If both directories exist, get list of signal and background files
+    if (
+        Path(config.signal_parent_dir).exists()
+        and Path(config.background_parent_dir).exists()
+    ):
+        config.list_signal_files = [
+            f for f in Path(config.signal_parent_dir).iterdir() if f.is_file()
+        ]
+        config.list_background_files = [
+            f
+            for f in Path(config.background_parent_dir).iterdir()
+            if f.is_file()
+        ]
+
+    # If exactly one of the input data directories is missing, print error
+    elif (
+        Path(config.signal_parent_dir).exists()
+        or Path(config.background_parent_dir).exists()
+    ):
+        if not Path(config.signal_parent_dir).exists():
+            logging.error(
+                f"The directory {config.signal_parent_dir} does not exist"
+            )
+        else:
+            logging.error(
+                f"The directory {config.background_parent_dir} does not exist"
+            )
+
+    # If neither of them exist, retrieve data from GIN repository
+    else:
+        if (not config.data_url) or (not config.data_hash):
+            logging.error(
+                "Input data not found locally, and URL/hash to "
+                "GIN repository not provided"
+            )
+
+        else:
+            # get list of files in GIN archive with retrieve
+            list_files_archive = pooch.retrieve(
+                url=config.data_url,
+                known_hash=config.data_hash,
+                path=config.install_path,  # path to download zip to
+                progressbar=True,
+                processor=pooch.Unzip(
+                    extract_dir=config.local_path  # path to unzipped dir
+                ),
+            )
+            logging.info(
+                "Fetching input data from the provided GIN repository"
+            )
+
+            # check signal and background parent directories exist now
+            assert Path(config.signal_parent_dir).exists()
+            assert Path(config.background_parent_dir).exists()
+
+            # add signal files to config
+            config.list_signal_files = [
+                f
+                for f in list_files_archive
+                if f.startswith(config.signal_parent_dir)
+            ]
+
+            # add background files to config
+            config.list_background_files = [
+                f
+                for f in list_files_archive
+                if f.startswith(config.background_parent_dir)
+            ]
+
+    return config
+
+
+if __name__ == "__main__":
+    example_cellfinder_script()

pyproject.toml

@@ -5,7 +5,10 @@ description = "A place to keep scripts to use as benchmarks, user-examples end-t
 readme = "README.md"
 requires-python = ">=3.8.0"
 dynamic = ["version"]
-
+dependencies = [
+    "pooch",
+    "cellfinder-core"
+]
 license = {text = "BSD-3-Clause"}
 
 classifiers = [
@@ -51,11 +54,11 @@ build-backend = "setuptools.build_meta"
 include-package-data = true
 
 [tool.setuptools.packages.find]
-include = ["brainglobe_scripts*"]
+include = ["brainglobe_workflows*"]
 exclude = ["tests*"]
 
 [tool.pytest.ini_options]
-addopts = "--cov=brainglobe_scripts"
+addopts = "--cov=brainglobe_workflows"
 
 [tool.black]
 target-version = ['py38', 'py39', 'py310']
@@ -101,5 +104,5 @@ python =
 extras =
     dev
 commands =
-    pytest -v --color=yes --cov=brainglobe_scripts --cov-report=xml
+    pytest -v --color=yes --cov=brainglobe_workflows --cov-report=xml
 """