Implement a working native pipeline

config.toml

@@ -0,0 +1,10 @@
+# Note: The sub-headings / sections are optional and are removed from the config parser anyway. The only important thing is the [darts] heading
+[darts.paths]
+input-data-dir = "data/input"
+output-data-dir = "data/output"
+model-dir = "models"
+
+[darts.segmentation]
+patch-size = 1024
+overlap = 128
+batch-size = 2

darts-export/src/darts_export/__init__.py

@@ -1 +1,3 @@
-# noqa: D104
+"""Dataset export for the DARTS dataset."""
+
+from darts_export.inference import InferenceResultWriter as InferenceResultWriter

darts-postprocessing/src/darts_postprocessing/__init__.py

@@ -1,2 +1,3 @@
-def hello() -> str:
-    return "Hello from darts-postprocessing!"
+"""Postprocessing steps for the DARTS dataset."""
+
+from darts_postprocessing.prepare_export import prepare_export as prepare_export

darts-preprocessing/src/darts_preprocessing/__init__.py

@@ -1,11 +1,3 @@
 """Data preprocessing and feature engineering for the DARTS dataset."""
 
-
-def hello() -> str:
-    """Say hello to the user.
-
-    Returns:
-        str: Greating message.
-
-    """
-    return "Hello from darts-preprocessing!"
+from darts_preprocessing.preprocess import load_and_preprocess_planet_scene as load_and_preprocess_planet_scene

darts-preprocessing/src/darts_preprocessing/data_sources/planet.py

@@ -51,3 +51,55 @@ def load_planet_scene(fpath: str | Path) -> xr.Dataset:
     ds_planet = xr.merge(datasets)
     logger.debug(f"Loaded Planet scene in {time.time() - start_time} seconds.")
     return ds_planet
+
+
+def load_planet_masks(fpath: str | Path) -> xr.Dataset:
+    """Load the valid and quality data masks from a Planet scene.
+
+    Args:
+        fpath (str | Path): The file path to the Planet scene from which to derive the masks.
+
+    Raises:
+        FileNotFoundError: If no matching UDM-2 TIFF file is found in the specified path.
+
+    Returns:
+        xr.Dataset: A merged xarray Dataset containing two data masks:
+            - 'valid_data_mask': A mask indicating valid (1) and no data (0).
+            - 'quality_data_mask': A mask indicating high quality (1) and low quality (0).
+
+    Notes:
+        - The function utilizes the `get_planet_udm2path_from_planet_scene_path`
+          to obtain the path to the UDM (User Data Model) file.
+        - It uses `rasterio` to read the UDM file and `xarray` for handling
+
+    """
+    start_time = time.time()
+    logger.debug(f"Loading data masks from {fpath}")
+    # Convert to Path object if a string is provided
+    fpath = fpath if isinstance(fpath, str) else Path(fpath)
+
+    # Get imagepath
+    udm_path = next(fpath.glob("*_udm2.tif"))
+    if not udm_path:
+        raise FileNotFoundError(f"No matching UDM-2 TIFF files found in {fpath} (.glob('*_udm2.tif'))")
+
+    da_udm = xr.open_dataarray(udm_path)
+
+    # valid data mask: valid data = 1, no data = 0
+    valid_data_mask = (
+        (da_udm.sel(band=8) == 0)
+        .assign_attrs({"data_source": "planet", "long_name": "Valid data mask"})
+        .to_dataset(name="valid_data_mask")
+        .drop_vars("band")
+    )
+
+    # quality data mask: high quality = 1, low quality = 0
+    quality_data_mask = (
+        (da_udm.sel(band=[2, 3, 4, 5, 6]).max(axis=0) != 1)
+        .assign_attrs({"data_source": "planet", "long_name": "Quality data mask"})
+        .to_dataset(name="quality_data_mask")
+    )
+
+    qa_ds = xr.merge([valid_data_mask, quality_data_mask])
+    logger.debug(f"Loaded data masks in {time.time() - start_time} seconds.")
+    return qa_ds