Some fixes for the DLC module

CHANGELOG.md

@@ -1,5 +1,6 @@
 # Upcoming
 
+* Fixed bug with empty `NWBInterface` out of `DeeplabcutInterface` conversion; sped up `DeeplabcutInterface` conversion when timestamps are specified. [PR #967](https://github.com/catalystneuro/neuroconv/pull/967)
 
 ## v0.5.0 (July 17, 2024)
 

requirements-testing.txt

@@ -6,3 +6,4 @@ ndx-miniscope
 spikeinterface[qualitymetrics]>=0.100.0
 zarr<2.18.0  # Error with Blosc (read-only during decode) in numcodecs on May 7; check later if resolved
 pytest-xdist
+ndx-pose

src/neuroconv/datainterfaces/behavior/deeplabcut/_dlc_utils.py

@@ -1,5 +1,4 @@
 import importlib
-import os
 import pickle
 import warnings
 from pathlib import Path
@@ -8,6 +7,7 @@
 import numpy as np
 import pandas as pd
 import yaml
+from ndx_pose import PoseEstimation, PoseEstimationSeries
 from pynwb import NWBFile
 from ruamel.yaml import YAML
 
@@ -20,11 +20,11 @@ def _read_config(config_file_path):
     """
     ruamelFile = YAML()
     path = Path(config_file_path)
-    if os.path.exists(path):
+    if path.exists():
         try:
             with open(path, "r") as f:
                 cfg = ruamelFile.load(f)
-                curr_dir = os.path.dirname(config_file_path)
+                curr_dir = config_file_path.parent
                 if cfg["project_path"] != curr_dir:
                     cfg["project_path"] = curr_dir
         except Exception as err:
@@ -58,14 +58,11 @@ def _get_movie_timestamps(movie_file, VARIABILITYBOUND=1000, infer_timestamps=Tr
     n_frames = int(reader.get(cv2.CAP_PROP_FRAME_COUNT))
     fps = reader.get(cv2.CAP_PROP_FPS)
 
+    warnings.warn("Inferring timestamps from video. This might take a while (to speed up, set timestamps)")
     for _ in range(n_frames):
         _ = reader.read()
         timestamps.append(reader.get(cv2.CAP_PROP_POS_MSEC))
 
-    for _ in range(len(reader)):
-        _ = reader.read()
-        timestamps.append(reader.get(cv2.CAP_PROP_POS_MSEC))
-
     timestamps = np.array(timestamps) / 1000  # Convert to seconds
 
     if np.nanvar(np.diff(timestamps)) < 1.0 / fps * 1.0 / VARIABILITYBOUND:
@@ -110,39 +107,51 @@ def _infer_nan_timestamps(timestamps):
     return timestamps
 
 
-def _ensure_individuals_in_header(df, dummy_name):
+def _ensure_individuals_in_header(df, individual_name):
     if "individuals" not in df.columns.names:
         # Single animal project -> add individual row to
         # the header of single animal projects.
-        temp = pd.concat({dummy_name: df}, names=["individuals"], axis=1)
+        temp = pd.concat({individual_name: df}, names=["individuals"], axis=1)
         df = temp.reorder_levels(["scorer", "individuals", "bodyparts", "coords"], axis=1)
     return df
 
 
-def _get_pes_args(config_file, h5file, individual_name, infer_timestamps=True):
-    if "DLC" not in h5file or not h5file.endswith(".h5"):
+def _get_pes_args(
+    *,
+    config_file: Path,
+    h5file: Path,
+    individual_name: str,
+    timestamps_available: bool = False,
+    infer_timestamps: bool = True,
+):
+    config_file = Path(config_file)
+    h5file = Path(h5file)
+
+    if "DLC" not in h5file.name or not h5file.suffix == ".h5":
         raise IOError("The file passed in is not a DeepLabCut h5 data file.")
 
     cfg = _read_config(config_file)
 
-    vidname, scorer = os.path.split(h5file)[-1].split("DLC")
-    scorer = "DLC" + os.path.splitext(scorer)[0]
+    vidname, scorer = h5file.stem.split("DLC")
+    scorer = "DLC" + scorer
     video = None
 
     df = _ensure_individuals_in_header(pd.read_hdf(h5file), individual_name)
 
     # Fetch the corresponding metadata pickle file
     paf_graph = []
-    filename, _ = os.path.splitext(h5file)
+    filename = str(h5file.parent / h5file.stem)
     for i, c in enumerate(filename[::-1]):
         if c.isnumeric():
             break
     if i > 0:
         filename = filename[:-i]
-    metadata_file = filename + "_meta.pickle"
-    if os.path.isfile(metadata_file):
+    metadata_file = Path(filename + "_meta.pickle")
+
+    if metadata_file.exists():
         with open(metadata_file, "rb") as file:
             metadata = pickle.load(file)
+
         test_cfg = metadata["data"]["DLC-model-config file"]
         paf_graph = test_cfg.get("partaffinityfield_graph", [])
         if paf_graph:
@@ -157,13 +166,19 @@ def _get_pes_args(config_file, h5file, individual_name, infer_timestamps=True):
             video = video_path, params["crop"]
             break
 
+    # find timestamps only if required:
+    if timestamps_available:
+        timestamps = None
+    else:
+        if video is None:
+            timestamps = df.index.tolist()  # setting timestamps to dummy TODO: extract timestamps in DLC?
+        else:
+            timestamps = _get_movie_timestamps(video[0], infer_timestamps=infer_timestamps)
+
     if video is None:
         warnings.warn(f"The video file corresponding to {h5file} could not be found...")
         video = "fake_path", "0, 0, 0, 0"
 
-        timestamps = df.index.tolist()  # setting timestamps to dummy TODO: extract timestamps in DLC?
-    else:
-        timestamps = _get_movie_timestamps(video[0], infer_timestamps=infer_timestamps)
     return scorer, df, video, paf_graph, timestamps, cfg
 
 
@@ -178,13 +193,11 @@ def _write_pes_to_nwbfile(
     exclude_nans,
     pose_estimation_container_kwargs: Optional[dict] = None,
 ):
-    from ndx_pose import PoseEstimation, PoseEstimationSeries
-
     pose_estimation_container_kwargs = pose_estimation_container_kwargs or dict()
 
     pose_estimation_series = []
-    for kpt, xyp in df_animal.groupby(level="bodyparts", axis=1, sort=False):
-        data = xyp.to_numpy()
+    for keypoint in df_animal.columns.get_level_values("bodyparts").unique():
+        data = df_animal.xs(keypoint, level="bodyparts", axis=1).to_numpy()
 
         if exclude_nans:
             # exclude_nans is inverse infer_timestamps. if not infer, there may be nans
@@ -194,8 +207,8 @@ def _write_pes_to_nwbfile(
             timestamps_cleaned = timestamps
 
         pes = PoseEstimationSeries(
-            name=f"{animal}_{kpt}",
-            description=f"Keypoint {kpt} from individual {animal}.",
+            name=f"{animal}_{keypoint}" if animal else keypoint,
+            description=f"Keypoint {keypoint} from individual {animal}.",
             data=data[:, :2],
             unit="pixels",
             reference_frame="(0,0) corresponds to the bottom left corner of the video.",
@@ -269,11 +282,17 @@ def add_subject_to_nwbfile(
     nwbfile : pynwb.NWBFile
         nwbfile with pes written in the behavior module
     """
-    scorer, df, video, paf_graph, dlc_timestamps, _ = _get_pes_args(config_file, h5file, individual_name)
+    timestamps_available = timestamps is not None
+    scorer, df, video, paf_graph, dlc_timestamps, _ = _get_pes_args(
+        config_file=config_file,
+        h5file=h5file,
+        individual_name=individual_name,
+        timestamps_available=timestamps_available,
+    )
     if timestamps is None:
         timestamps = dlc_timestamps
 
-    df_animal = df.groupby(level="individuals", axis=1).get_group(individual_name)
+    df_animal = df.xs(individual_name, level="individuals", axis=1)
 
     return _write_pes_to_nwbfile(
         nwbfile,

src/neuroconv/datainterfaces/behavior/deeplabcut/deeplabcutdatainterface.py

@@ -4,6 +4,7 @@
 import numpy as np
 from pynwb.file import NWBFile
 
+from ._dlc_utils import add_subject_to_nwbfile
 from ....basetemporalalignmentinterface import BaseTemporalAlignmentInterface
 from ....utils import FilePathType
 
@@ -103,7 +104,6 @@ def add_to_nwbfile(
         metadata: dict
             metadata info for constructing the nwb file (optional).
         """
-        from ._dlc_utils import add_subject_to_nwbfile
 
         add_subject_to_nwbfile(
             nwbfile=nwbfile,