actinet 0.7.2__tar.gz → 0.8.0__tar.gz

{actinet-0.7.2 → actinet-0.8.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: actinet
-Version: 0.7.2
+Version: 0.8.0
 Summary: Activity detection algorithm compatible with the UK Biobank Accelerometer Dataset
 Home-page: https://github.com/OxWearables/actinet
 Download-URL: https://github.com/OxWearables/actinet
@@ -34,10 +34,14 @@ Requires-Dist: versioneer; extra == "dev"
 Requires-Dist: twine; extra == "dev"
 Requires-Dist: ipython; extra == "dev"
 Requires-Dist: ipdb; extra == "dev"
-Requires-Dist: flake8; extra == "dev"
+Requires-Dist: ruff==0.15.20; extra == "dev"
 Requires-Dist: autopep8; extra == "dev"
 Requires-Dist: tomli; extra == "dev"
 Requires-Dist: jupyter; extra == "dev"
+Requires-Dist: mypy==1.14.1; extra == "dev"
+Requires-Dist: pytest<8.4,>=8.3; extra == "dev"
+Requires-Dist: pytest-cov<6,>=5; extra == "dev"
+Requires-Dist: seaborn==0.13.*; extra == "dev"
 Provides-Extra: docs
 Requires-Dist: sphinx>=4.2; extra == "docs"
 Requires-Dist: sphinx_rtd_theme>=1.0; extra == "docs"
@@ -117,6 +121,9 @@ $ actinet sample.csv --csv-txyz 'timestamp,accel_x,accel_y,accel_z'
 
 # Or specify columns by index (0-indexed)
 $ actinet sample.csv --csv-txyz-idxs '0,1,2,3'
+
+# Process an inclusive time interval from the recording
+$ actinet sample.cwa --start '2024-01-01 10:00:00' --end '2024-01-08 09:59:59'
 ```
 
 See the [Usage](https://actinet.readthedocs.io/en/latest/usage.html) page for further uses of the tool.

{actinet-0.7.2 → actinet-0.8.0}/README.md

@@ -56,6 +56,9 @@ $ actinet sample.csv --csv-txyz 'timestamp,accel_x,accel_y,accel_z'
 
 # Or specify columns by index (0-indexed)
 $ actinet sample.csv --csv-txyz-idxs '0,1,2,3'
+
+# Process an inclusive time interval from the recording
+$ actinet sample.cwa --start '2024-01-01 10:00:00' --end '2024-01-08 09:59:59'
 ```
 
 See the [Usage](https://actinet.readthedocs.io/en/latest/usage.html) page for further uses of the tool.

{actinet-0.7.2 → actinet-0.8.0}/pyproject.toml

@@ -116,7 +116,7 @@
 # # Similar to `dependencies` above, these must be valid existing
 # # projects.
 # [project.optional-dependencies] # Optional
-# dev = ["check-manifest", "build", "twine", "versioneer", "flake8", "autopep8", "ipdb"]
+# dev = ["check-manifest", "build", "twine", "versioneer", "ruff==0.15.20", "autopep8", "ipdb"]
 # test = ["coverage"]
 
 # # List URLs that are relevant to your project
@@ -161,3 +161,72 @@ versionfile_source = "src/actinet/_version.py"
 versionfile_build = "actinet/_version.py"
 tag_prefix = "v"
 parentdir_prefix = "actinet-"
+
+[tool.ruff]
+target-version = "py38"
+line-length = 120
+extend-exclude = ["venv", "versioneer.py", "src/actinet/_version.py"]
+
+[tool.ruff.lint]
+select = [
+    "E4",   # pycodestyle imports
+    "E7",   # pycodestyle statements
+    "E9",   # pycodestyle runtime errors
+    "F",    # Pyflakes
+    "B",    # flake8-bugbear
+    "C4",   # flake8-comprehensions
+    "C90",  # mccabe
+    "I",    # isort
+]
+
+[tool.ruff.lint.mccabe]
+max-complexity = 10
+
+[tool.ruff.lint.per-file-ignores]
+"__init__.py" = ["F401"]
+
+[tool.mypy]
+python_version = "3.8"
+files = ["src/actinet"]
+exclude = [
+    "^src/actinet/_version\\.py$",
+    "^src/actinet/torch_hub_cache/",
+]
+check_untyped_defs = true
+disallow_untyped_defs = true
+no_implicit_optional = true
+show_error_codes = true
+warn_redundant_casts = true
+warn_unused_ignores = true
+
+[[tool.mypy.overrides]]
+module = [
+    "actipy.*",
+    "imblearn.*",
+    "joblib.*",
+    "matplotlib.*",
+    "pandas.*",
+    "scipy.*",
+    "seaborn.*",
+    "sklearn.*",
+    "tqdm.*",
+    "transforms3d.*",
+    "torchvision.*",
+]
+ignore_missing_imports = true
+
+[[tool.mypy.overrides]]
+module = "actinet._version"
+follow_imports = "skip"
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+markers = [
+  "slow: real-device or otherwise longer-running regression tests",
+]
+
+[tool.coverage.run]
+omit = ["*/actinet/_version.py"]
+
+[tool.coverage.report]
+fail_under = 80

{actinet-0.7.2 → actinet-0.8.0}/setup.py

@@ -46,7 +46,7 @@ def main():
         packages=find_packages(where="src", exclude=("test", "tests")),
         package_dir={"": "src"},
         # What other files to include, e.g. *.class if the package uses some Java code.
-        package_data={"actinet": ["*.txt", "*.rst", "*.md"]},
+        package_data={"actinet": ["*.txt", "*.rst", "*.md", "py.typed"]},
         include_package_data=False,
         # Dependencies
         install_requires=[
@@ -69,10 +69,14 @@ def main():
                 "twine",
                 "ipython",
                 "ipdb",
-                "flake8",
+                "ruff==0.15.20",
                 "autopep8",
                 "tomli",
                 "jupyter",
+                "mypy==1.14.1",
+                "pytest>=8.3,<8.4",
+                "pytest-cov>=5,<6",
+                "seaborn==0.13.*",
             ],
             "docs": [
                 "sphinx>=4.2",

{actinet-0.7.2 → actinet-0.8.0}/src/actinet/_version.py

@@ -8,11 +8,11 @@ import json
 
 version_json = '''
 {
- "date": "2026-03-31T10:30:51+0100",
+ "date": "2026-10-08T18:41:48+0100",
  "dirty": false,
  "error": null,
- "full-revisionid": "58d134b8be0965b5611aa589bc07a926de68f575",
- "version": "0.7.2"
+ "full-revisionid": "497eea3d3834609c45be709c9a70cff0a59235ea",
+ "version": "0.8.0"
 }
 '''  # END VERSION_JSON
 

{actinet-0.7.2 → actinet-0.8.0}/src/actinet/accPlot.py

@@ -1,6 +1,8 @@
 """Script to plot accelerometer traces."""
 
 import sys
+from typing import Any, Optional
+
 import numpy as np
 import pandas as pd
 from pandas.plotting import register_matplotlib_converters
@@ -39,7 +41,7 @@ LABELS_AND_COLORS = {
 }
 
 
-def main():  # noqa: C901
+def main() -> None:  # noqa: C901
     """
     Application entry point responsible for parsing command line requests
     """
@@ -111,7 +113,11 @@ def main():  # noqa: C901
     print("Plot file written to:", args.plotFile)
 
 
-def plotTimeSeries(data, title=None, showFirstNDays=None):  # noqa: C901
+def plotTimeSeries(
+    data: pd.DataFrame,
+    title: Optional[str] = None,
+    showFirstNDays: Optional[int] = None,
+) -> Any:  # noqa: C901
     """
     Plot acceleration traces and classified activities.
 
@@ -292,7 +298,7 @@ def plotTimeSeries(data, title=None, showFirstNDays=None):  # noqa: C901
     return fig
 
 
-def str2bool(v):
+def str2bool(v: str) -> bool:
     """
     Used to parse true/false values from the command line. E.g. "True" -> True
     """

{actinet-0.7.2 → actinet-0.8.0}/src/actinet/actinet.py

@@ -1,36 +1,39 @@
+from __future__ import annotations
+
+import argparse
+import hashlib
+import json
 import os
 import pathlib
-import urllib
 import shutil
+import sys
 import time
-import argparse
-import json
-import hashlib
+import urllib
 import warnings
-import numpy as np
-import pandas as pd
-import joblib
-import sys
+from typing import TYPE_CHECKING, Any, Dict, Literal, Optional, Tuple, Union
 
-import actipy
-
-from actinet import __version__
-from actinet import __classifiers__
-from actinet.accPlot import plotTimeSeries
-from actinet.models import ActivityClassifier
-from actinet.summarisation import get_activity_summary
-from actinet.utils.summary_utils import calculate_daily_wear_stats
-from actinet.utils.utils import (
-    infer_freq,
-    drop_first_last_days,
-    flag_wear_below_days,
-    calculate_wear_stats,
-)
+from actinet import __classifiers__, __version__
+
+if TYPE_CHECKING:
+    import pandas as pd
+
+    from actinet.models import ActivityClassifier
+
+
+def __getattr__(name: str) -> Any:
+    if name == "ActivityClassifier":
+        from actinet.models import ActivityClassifier
+
+        globals()[name] = ActivityClassifier
+        return ActivityClassifier
+    raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
 
 BASE_URL = "https://wearables-files.ndph.ox.ac.uk/files/models/actinet/"
 
 
-def save_and_print_summary(outputSummaryFile, info, verbose):
+def save_and_print_summary(
+    outputSummaryFile: str, info: Dict[str, Any], verbose: bool
+) -> None:
     """Save info to outputSummary.json and print summary stats."""
     with open(outputSummaryFile, "w") as f:
         json.dump(info, f, indent=4, cls=NpEncoder)
@@ -55,7 +58,7 @@ def save_and_print_summary(outputSummaryFile, info, verbose):
         )
 
 
-def main():
+def main() -> None:
     parser = argparse.ArgumentParser(
         description="A tool to predict activities from accelerometer data using a self-supervised ResNet-18 model",
         add_help=True,
@@ -126,6 +129,24 @@ def main():
         type=str,
         default=None,
     )
+    parser.add_argument(
+        "--start",
+        help=(
+            "Specify a start time for the data to be processed (otherwise, process all). "
+            "Pass values as strings, e.g.: '2024-01-01 10:00:00'. Default: None"
+        ),
+        type=str,
+        default=None,
+    )
+    parser.add_argument(
+        "--end",
+        help=(
+            "Specify an end time for the data to be processed (otherwise, process all). "
+            "Pass values as strings, e.g.: '2024-01-02 09:59:59'. Default: None"
+        ),
+        type=str,
+        default=None,
+    )
     parser.add_argument(
         "--csv-start-row",
         help="Row number to start reading a CSV file. Default: 1 (First row)",
@@ -168,7 +189,7 @@ def main():
     parser.add_argument(
         "--cache-classifier",
         action="store_true",
-        help="Download and cache classifier file and model modules for offline usage",
+        help="Download and cache classifier file and model modules for offline usage. This will cause the ActiNet classification model to not run.",
     )
     parser.add_argument(
         "--model-repo-path", "-m", help="Enter repository of ssl model", default=None
@@ -197,6 +218,13 @@ def main():
         if not args.filepath:
             raise ValueError("Please provide a file to process.")
 
+    from actinet.utils.summary_utils import calculate_daily_wear_stats
+    from actinet.utils.utils import (
+        calculate_wear_stats,
+        drop_first_last_days,
+        flag_wear_below_days,
+    )
+
     # Info contains high-level summary of the data and results
     info = {}
     info["ActiNetVersion"] = __version__
@@ -211,6 +239,8 @@ def main():
         args.calibration_stdtol_min,
         resample_hz="uniform",
         sample_rate=args.sample_rate,
+        start_time=args.start,
+        end_time=args.end,
         csv_txyz_idxs=args.csv_txyz_idxs,
         verbose=verbose,
     )
@@ -275,6 +305,11 @@ def main():
             args.require_sleep_above,
             args.single_sleep_block,
         )
+        if len(Y) < 3:
+            raise ValueError(
+                "The selected time interval must produce at least three "
+                f"{classifier.window_sec}-second activity epochs."
+            )
     except Exception as e:
         save_and_print_summary(outputSummaryFile, info, verbose)
         raise ValueError(f"Model failed to make predictions on the data: {str(e)}")
@@ -290,6 +325,8 @@ def main():
 
     # Plot activity profile
     if args.plot_activity:
+        from actinet.accPlot import plotTimeSeries
+
         plotFile = f"{outdir}/{basename}-timeSeries-plot.png"
         fig = plotTimeSeries(Y)
         fig.savefig(plotFile, dpi=200, bbox_inches="tight")
@@ -298,6 +335,8 @@ def main():
             print("Output plot written to:", plotFile)
 
     # Summary
+    from actinet.summarisation import get_activity_summary
+
     summary, daily_summary = get_activity_summary(
         Y, list(classifier.labels), args.exclude_wear_below, True, True, verbose
     )
@@ -312,6 +351,8 @@ def main():
     if verbose:
         print("Output summary written to:", outputSummaryFile)
 
+    import pandas as pd
+
     daily_summary = pd.concat([daily_wear_stats, daily_summary], axis=1)
 
     daily_summary.insert(0, "Filename", info["Filename"])  # add filename for reference
@@ -346,17 +387,27 @@ def main():
 
 
 def read(
-    filepath,
-    usecols=None,
-    skipRows=0,
-    dateFormat=None,
-    calibration_stdtol_min=None,
-    resample_hz="uniform",
-    sample_rate=None,
-    lowpass_hz=None,
-    csv_txyz_idxs=None,
-    verbose=True,
-):
+    filepath: str,
+    usecols: Optional[str] = None,
+    skipRows: int = 0,
+    dateFormat: Optional[str] = None,
+    calibration_stdtol_min: Optional[float] = None,
+    resample_hz: Optional[Union[Literal["uniform"], int, float, bool]] = "uniform",
+    sample_rate: Optional[int] = None,
+    lowpass_hz: Optional[float] = None,
+    csv_txyz_idxs: Optional[str] = None,
+    verbose: bool = True,
+    start_time: Any = None,
+    end_time: Any = None,
+) -> Tuple[pd.DataFrame, Dict[str, Any]]:
+    import actipy
+    import numpy as np
+    import pandas as pd
+
+    from actinet.utils.utils import infer_freq
+
+    start_time, end_time = validate_time_interval(start_time, end_time)
+
     p = pathlib.Path(filepath)
     fsize = round(p.stat().st_size / (1024 * 1024), 1)
 
@@ -397,6 +448,8 @@ def read(
                     header[zidx],
                 )
             else:
+                if usecols is None:
+                    usecols = "time,x,y,z"
                 tcol, xcol, ycol, zcol = usecols.split(",")
 
             data = pd.read_csv(
@@ -420,6 +473,8 @@ def read(
             freq = infer_freq(data.index)
             sample_rate = int(np.round(pd.Timedelta("1s") / freq))
 
+        assert sample_rate is not None
+
         # Quick fix: Drop duplicate indices. TODO: Maybe should be handled by actipy.
         data = data[~data.index.duplicated(keep="first")]
 
@@ -466,10 +521,77 @@ def read(
     if "ResampleRate" not in info:
         info["ResampleRate"] = info["SampleRate"]
 
+    start_time, end_time = validate_time_interval(
+        start_time, end_time, data.index
+    )
+
+    # Apply bounds after ActiPy processes the complete recording for calibration
+    # and non-wear detection.
+    if start_time is not None:
+        data = data.loc[start_time:]
+    if end_time is not None:
+        data = data.loc[:end_time]
+
+    if len(data) == 0 and (start_time is not None or end_time is not None):
+        raise ValueError("The selected time interval does not contain any data.")
+
     return data, info
 
 
-def resolve_path(path):
+def validate_time_interval(
+    start_time: Any = None,
+    end_time: Any = None,
+    index: Optional[pd.Index] = None,
+) -> Tuple[Optional[pd.Timestamp], Optional[pd.Timestamp]]:
+    """Parse optional bounds and validate them against a recording index."""
+
+    import pandas as pd
+
+    def parse_bound(value: Any, name: str) -> Optional[pd.Timestamp]:
+        if value is None:
+            return None
+        try:
+            value = pd.Timestamp(value)
+        except (TypeError, ValueError) as error:
+            raise ValueError(f"Invalid {name} time: {value!r}") from error
+        if pd.isna(value):
+            raise ValueError(f"Invalid {name} time: {value!r}")
+        return value
+
+    start_time = parse_bound(start_time, "start")
+    end_time = parse_bound(end_time, "end")
+
+    if start_time is not None and end_time is not None:
+        try:
+            reversed_interval = start_time > end_time
+        except TypeError as error:
+            raise ValueError(
+                "Start and end times must use compatible timezones."
+            ) from error
+        if reversed_interval:
+            raise ValueError("Start time must be earlier than or equal to end time.")
+
+    if index is None or len(index) == 0:
+        return start_time, end_time
+    if not isinstance(index, pd.DatetimeIndex):
+        raise ValueError("Time interval selection requires a DatetimeIndex.")
+
+    index_is_aware = index.tz is not None
+    for name, bound in (("start", start_time), ("end", end_time)):
+        if bound is not None and (bound.tzinfo is not None) != index_is_aware:
+            raise ValueError(
+                f"The {name} time and recording timestamps must use compatible timezones."
+            )
+
+    if start_time is not None and start_time > index[-1]:
+        raise ValueError("The selected time interval does not overlap the recording.")
+    if end_time is not None and end_time < index[0]:
+        raise ValueError("The selected time interval does not overlap the recording.")
+
+    return start_time, end_time
+
+
+def resolve_path(path: Union[str, pathlib.Path]) -> Tuple[pathlib.Path, str, str]:
     """Return parent folder, file name and file extension"""
     p = pathlib.Path(path)
     extension = p.suffixes[0]
@@ -479,13 +601,15 @@ def resolve_path(path):
 
 
 def load_classifier(
-    classifier,
-    model_repo_path=None,
-    force_download=False,
-    verbose=True,
-):
+    classifier: str,
+    model_repo_path: Optional[str] = None,
+    force_download: bool = False,
+    verbose: bool = True,
+) -> ActivityClassifier:
     """Load trained classifier. Download if not exists."""
 
+    import joblib
+
     if classifier in __classifiers__.keys():
         classifier_version = __classifiers__[classifier]["version"]
         classifier_md5 = __classifiers__[classifier]["md5"]
@@ -514,25 +638,25 @@ def load_classifier(
                     )
 
     elif pathlib.Path(classifier).exists():
-        classifier_path = classifier
+        classifier_path = pathlib.Path(classifier)
 
     else:
         raise ValueError(f"Unknown classifier: {classifier}")
 
     try:
-        classifier: ActivityClassifier = joblib.load(classifier_path)
+        loaded_classifier: ActivityClassifier = joblib.load(classifier_path)
     except Exception as e:
-        raise ValueError(f"Error loading classifier file.")
+        raise ValueError("Error loading classifier file.")
 
     if model_repo_path and pathlib.Path(model_repo_path).exists() and verbose:
         print(f"Loading model repository from {model_repo_path}.")
 
-    classifier.load_model(model_repo_path)
+    loaded_classifier.load_model(model_repo_path)
 
-    return classifier
+    return loaded_classifier
 
 
-def md5(fname):
+def md5(fname: Union[str, pathlib.Path]) -> str:
     hash_md5 = hashlib.md5()
     with open(fname, "rb") as f:
         for chunk in iter(lambda: f.read(4096), b""):
@@ -541,7 +665,12 @@ def md5(fname):
 
 
 class NpEncoder(json.JSONEncoder):
-    def default(self, obj):
+    def default(self, obj: Any) -> Any:
+        import numpy as np
+        import pandas as pd
+
+        if isinstance(obj, np.bool_):
+            return bool(obj)
         if isinstance(obj, np.integer):
             return int(obj)
         if isinstance(obj, np.floating):