FIX : EDF+ Annotation Timestamps missing sub-second accuracy (#7875)

* FIX : EDF+ Annotation Timestamps missing submillisecond accuracy * review from @cbrnr * udpate test + testing data * rephrase * Rename data Co-authored-by: Clemens Brunner <[email protected]>
mne-tools · Jun 8, 2020 · b2615a3 · b2615a3
1 parent f8d2ce1
commit b2615a3
Show file tree

Hide file tree

Showing 3 changed files with 23 additions and 4 deletions.
diff --git a/mne/datasets/utils.py b/mne/datasets/utils.py
@@ -239,7 +239,7 @@ def _data_path(path=None, force_update=False, update_path=True, download=True,
     path = _get_path(path, key, name)
     # To update the testing or misc dataset, push commits, then make a new
     # release on GitHub. Then update the "releases" variable:
-    releases = dict(testing='0.91', misc='0.6')
+    releases = dict(testing='0.92', misc='0.6')
     # And also update the "md5_hashes['testing']" variable below.
 
     # To update any other dataset, update the data archive itself (upload
@@ -326,7 +326,7 @@ def _data_path(path=None, force_update=False, update_path=True, download=True,
         sample='12b75d1cb7df9dfb4ad73ed82f61094f',
         somato='ea825966c0a1e9b2f84e3826c5500161',
         spm='9f43f67150e3b694b523a21eb929ea75',
-        testing='f87f04fcdf56a7a55fb73b1b260b5b5b',
+        testing='42daafd1b882da2ef041de860ca6e771',
         multimodal='26ec847ae9ab80f58f204d09e2c08367',
         fnirs_motor='c4935d19ddab35422a69f3326a01fef8',
         opm='370ad1dcfd5c47e029e692c85358a374',

diff --git a/mne/io/edf/edf.py b/mne/io/edf/edf.py
@@ -1375,12 +1375,23 @@ def _read_annotations_edf(annotations):
         triggers = re.findall(pat, tals.decode('latin-1'))
 
     events = []
-    for ev in triggers:
-        onset = float(ev[0])
+    offset = 0.
+    for k, ev in enumerate(triggers):
+        onset = float(ev[0]) + offset
         duration = float(ev[2]) if ev[2] else 0
         for description in ev[3].split('\x14')[1:]:
             if description:
                 events.append([onset, duration, description])
+            elif k == 0:
+                # The startdate/time of a file is specified in the EDF+ header
+                # fields 'startdate of recording' and 'starttime of recording'.
+                # These fields must indicate the absolute second in which the
+                # start of the first data record falls. So, the first TAL in
+                # the first data record always starts with +0.X, indicating
+                # that the first data record starts a fraction, X, of a second
+                # after the startdate/time that is specified in the EDF+
+                # header. If X=0, then the .X may be omitted.
+                offset = -onset
 
     return zip(*events) if events else (list(), list(), list())
 

diff --git a/mne/io/edf/tests/test_edf.py b/mne/io/edf/tests/test_edf.py
@@ -56,6 +56,7 @@
                                         'multiple_annotation_chans.bdf')
 test_generator_bdf = op.join(data_path, 'BDF', 'test_generator_2.bdf')
 test_generator_edf = op.join(data_path, 'EDF', 'test_generator_2.edf')
+edf_annot_sub_s_path = op.join(data_path, 'EDF', 'subsecond_starttime.edf')
 
 eog = ['REOG', 'LEOG', 'IEOG']
 misc = ['EXG1', 'EXG5', 'EXG8', 'M1', 'M2']
@@ -385,3 +386,10 @@ def test_edf_lowpass_zero():
     with pytest.warns(RuntimeWarning, match='too long.*truncated'):
         raw = read_raw_edf(edf_stim_resamp_path)
     assert_allclose(raw.info["lowpass"], raw.info["sfreq"] / 2)
+
+
+@testing.requires_testing_data
+def test_edf_annot_sub_s_onset():
+    """Test reading of sub-second annotation onsets."""
+    raw = read_raw_edf(edf_annot_sub_s_path)
+    assert_allclose(raw.annotations.onset, [1.951172, 3.492188])