✨ feat: Feature extraction with an identifier

pyproject.toml

@@ -79,9 +79,14 @@ ignore = ["E501"] # Never enforce `E501` (line length violations).
 "tests/test_stroll_factory.py" = ["F401", "F811"]
 "tests/test_utils.py" = ["F401", "F811"]
 
+# Testing
+[tool.pytest.ini_options]
+addopts = "--cov=tsflex --cov-report=term-missing"
+testpaths = "tests/"
+
 # Formatting
 [tool.black]
-color = true
+color = false
 line-length = 88
 
 [build-system]

tests/benchmarks/test_featurecollection.py

@@ -6,7 +6,7 @@
 from tsflex.features.feature import FeatureDescriptor, MultipleFeatureDescriptors
 from tsflex.features.feature_collection import FeatureCollection
 
-from ..utils import dummy_data  # noqa: F401
+from ..utils import dummy_data, dummy_group_data  # noqa: F401
 
 FUNCS = [np.sum, np.min, np.max, np.mean, np.median, np.std, np.var]
 MAX_CPUS = os.cpu_count() or 2
@@ -50,3 +50,17 @@ def test_single_series_feature_collection_multiple_descriptors(
     fc = FeatureCollection(mfd)
 
     benchmark(fc.calculate, dummy_data, n_jobs=n_cores)
+
+
+@pytest.mark.benchmark(group="group_by collection")
+@pytest.mark.parametrize("n_cores", NB_CORES)
+@pytest.mark.parametrize("func", FUNCS)
+@pytest.mark.parametrize("group_by", ["group_by_all", "group_by_consecutive"])
+def test_single_series_feature_collection_group_by_consecutive(
+    benchmark, n_cores, func, group_by, dummy_group_data  # noqa: F811
+):
+    fd = FeatureDescriptor(function=func, series_name="number_sold")
+
+    fc = FeatureCollection(feature_descriptors=fd)
+
+    benchmark(fc.calculate, dummy_group_data, n_jobs=n_cores, **{group_by: "store"})