Bump joblib from 1.1.0 to 1.2.0

.gitignore

@@ -126,4 +126,5 @@ doc/source/savefig/
 
 # Sample files, etc.
 data/
-accelerometer/activityModels/
+accelerometer/activityModels/
+accelerometer/models/

accelerometer/__init__.py

@@ -1,5 +1,5 @@
 name = "accelerometer"
-__version__ = "5.1.3"
+__version__ = "6.2.0"
 __author__ = "Aiden Doherty, Shing Chan, Rosemary Walmsley, Hang Yuan"
 __email__ = "aiden.doherty@ndph.ox.ac.uk, shing.chan@ndph.ox.ac.uk, rosemary.walmsley@gtc.ox.ac.uk, hang.yuan@keble.ox.ac.uk"
 __license__ = "See LICENSE.md"

accelerometer/accPlot.py

@@ -24,7 +24,7 @@
     'vehicle': 'saddlebrown',
     'light': 'darkorange',
     'mixed': 'seagreen',
-    'walking': 'green',
+    'walking': 'limegreen',
     'moderate-vigorous': 'green',
     'bicycling': 'springgreen',
     'tasks-light': 'darkorange',

accelerometer/classification.py

@@ -40,6 +40,9 @@ def activityClassification(epoch, activityModel="walmsley"):
     :rtype: list(str)
     """
 
+    use_cutpoints = 'chan' in activityModel
+    smooth_sleep = 'chan' in activityModel
+
     activityModel = resolveModelPath(activityModel)
 
     featureCols = joblib.load(getFileFromTar(activityModel, 'featureCols'))
@@ -51,8 +54,29 @@ def activityClassification(epoch, activityModel="walmsley"):
 
     model = joblib.load(getFileFromTar(activityModel, 'model'))
     hmmParams = joblib.load(getFileFromTar(activityModel, 'hmmParams'))
+    labels = joblib.load(getFileFromTar(activityModel, 'labels')).tolist()
+
     Y = viterbi(model.predict(X), hmmParams)
 
+    if smooth_sleep:
+        sleep = pd.Series(Y == 'sleep')
+        sleep_streak = (
+            sleep.ne(sleep.shift())
+            .cumsum()
+            .pipe(lambda x: x.groupby(x).transform('count') * sleep)
+        )
+        # TODO: hardcoded 120 = 1hr
+        Y[(Y == 'sleep') & (sleep_streak < 120)] = 'sedentary'
+
+    if use_cutpoints:
+        enmo = epoch['enmoTrunc'].to_numpy()
+        enmo = enmo[mask]
+        Y[(Y == 'other') & (enmo < .1)] = 'light'
+        Y[(Y == 'other') & (enmo >= .1)] = 'moderate-vigorous'
+        labels.remove('other')
+        labels.append('light')
+        labels.append('moderate-vigorous')
+
     # Append predicted activities to epoch dataframe
     epoch["label"] = np.nan
     epoch.loc[mask, "label"] = Y
@@ -62,8 +86,6 @@ def activityClassification(epoch, activityModel="walmsley"):
     if METs is not None:
         epoch["MET"] = epoch["label"].replace(METs)
 
-    labels = joblib.load(getFileFromTar(activityModel, 'labels')).tolist()
-
     # One-hot encoding
     for lab in labels:
         epoch[lab] = 0

accelerometer/models.py

@@ -1,36 +1,35 @@
 import pathlib
 
-ROOT_DIR = pathlib.Path(__file__).parent
-MODEL_VER = "10Feb2022"
-MODEL_DIR = ROOT_DIR / "models"
-MODEL_URL = "https://wearables-files.ndph.ox.ac.uk/files/models"
+MODEL_DIR = pathlib.Path(__file__).parent / "models"
+MODEL_ROOT_URL = "https://wearables-files.ndph.ox.ac.uk/files/models/biobankAccelerometerAnalysis"
 
 
 MODELS = {
+
     'willetts': {
-        "pth": MODEL_DIR / MODEL_VER / "willetts/model.tar",
-        "url": f"{MODEL_URL}/{MODEL_VER}/willetts/model.tar",
+        "pth": MODEL_DIR / "willetts" / "20220210.tar",
+        "url": f"{MODEL_ROOT_URL}/willetts/20220210.tar",
     },
+
     'doherty': {
-        "pth": MODEL_DIR / MODEL_VER / "doherty/model.tar",
-        "url": f"{MODEL_URL}/{MODEL_VER}/doherty/model.tar",
+        "pth": MODEL_DIR / "doherty" / "20220210.tar",
+        "url": f"{MODEL_ROOT_URL}/doherty/20220210.tar",
     },
+
     'walmsley': {
-        "pth": MODEL_DIR / MODEL_VER / "walmsley/model.tar",
-        "url": f"{MODEL_URL}/{MODEL_VER}/walmsley/model.tar",
+        "pth": MODEL_DIR / "walmsley" / "20220210.tar",
+        "url": f"{MODEL_ROOT_URL}/walmsley/20220210.tar",
     },
 
-    'willetts-10Feb2022': {
-        "pth": MODEL_DIR / "10Feb2022" / "willetts/model.tar",
-        "url": f"{MODEL_URL}/10Feb2022/walmsley/model.tar",
-    },
-    'doherty-10Feb2022': {
-        "pth": MODEL_DIR / "10Feb2022" / "doherty/model.tar",
-        "url": f"{MODEL_URL}/10Feb2022/walmsley/model.tar",
+    'chan': {
+        "pth": MODEL_DIR / "chan" / "20230103.tar",
+        "url": f"{MODEL_ROOT_URL}/chan/20230103.tar",
+
     },
-    'walmsley-10Feb2022': {
-        "pth": MODEL_DIR / "10Feb2022" / "walmsley/model.tar",
-        "url": f"{MODEL_URL}/10Feb2022/walmsley/model.tar",
+
+    'chanw': {
+        "pth": MODEL_DIR / "chanw" / "20230106.tar",
+        "url": f"{MODEL_ROOT_URL}/chanw/20230106.tar"
     },
 
 }

build_scripts/build_clean.sh

@@ -0,0 +1,4 @@
+#!/bin/bash
+
+rm -r actipy.egg-info build dist
+rm -r conda-recipe

build_scripts/build_conda.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+
+# Note: Be sure package already in PyPI
+
+conda install anaconda-client &&
+conda install conda-build &&
+# TODO: allow user-specified version and append this to next line: --version x.x.x
+conda skeleton pypi accelerometer --output-dir conda-recipe &&
+conda build -c conda-forge conda-recipe/accelerometer
+
+printf "\nNext steps:\n-----------\n"
+printf "Login to Anaconda:\n> anaconda login\n"
+printf "\nUpload package (path is printed in previous steps):\n> anaconda upload --user oxwear /path/to/package.tar.bz2\n\n"
+
+# anaconda login
+# anaconda upload --user oxwear /path/to/package.tar.bz2

build_scripts/build_pypi.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+
+# You must have Java Development Kit (JDK) 8 (1.8). If higher (>8) then it must
+# support --release flag to pin down the version when compiling.
+# Always compile with 8 (1.8) to keep backward compatibility.
+# In conda, you can get a JDK version that supports --release flag:
+# conda install openjdk
+javac --version &&  # java version
+javac -cp accelerometer/java/JTransforms-3.1-with-dependencies.jar accelerometer/java/*.java --release 8 &&  # compile java files (using release 8)
+python setup.py sdist bdist_wheel &&  # setuptools
+twine check dist/* &&
+printf "\nTo upload to Test PyPI:\n> twine upload --repository-url https://test.pypi.org/legacy/ dist/*\n" &&
+printf "\nTo upload to PyPI:\n> twine upload dist/*\n\n"
+
+# twine upload --repository-url https://test.pypi.org/legacy/ dist/*
+# twine upload dist/*

setup.py

@@ -39,9 +39,25 @@ def get_string(string, rel_path="accelerometer/__init__.py"):
         'statsmodels>=0.12.2',
         'imbalanced-learn==0.8.1',
         'scikit-learn==1.0.1',
-        'joblib==1.1.0',
+        'joblib==1.2.0',
         'tqdm>=4.59.0',
     ],
+    extras_require={
+        "dev": [
+            "flake8",
+            "autopep8",
+            "ipython",
+            "ipdb",
+            "twine",
+        ],
+        "docs": [
+            "sphinx>=4.2",
+            "sphinx_rtd_theme>=1.0",
+            "readthedocs-sphinx-search>=0.1",
+            "sphinxcontrib-programoutput>=0.17",
+            "docutils<0.18",
+        ],
+    },
     classifiers=[
         "Programming Language :: Python :: 3",
         "Operating System :: MacOS :: MacOS X",