ENH: Use date and datetime objects

doc/changes/devel/12720.bugfix.rst

@@ -0,0 +1 @@
+Fix in-memory anonymization of data read with :func:`mne.io.read_raw_edf` by `Eric Larson`_.

doc/changes/devel/12720.newfeature.rst

@@ -0,0 +1,2 @@
+Use :class:`python:datetime.date` for ``info["subject_info"]["birthday"]`` rather than
+a tuple of ``(year, month, day)`` by `Eric Larson`_.

mne/_fiff/meas_info.py

@@ -1021,6 +1021,13 @@ def _check_line_freq(line_freq, *, info):
 
 def _check_subject_info(subject_info, *, info):
     _validate_type(subject_info, (None, dict), "subject_info")
+    if isinstance(subject_info, dict):
+        if "birthday" in subject_info:
+            _validate_type(
+                subject_info["birthday"],
+                (datetime.date, None),
+                "subject_info['birthday']",
+            )
     return subject_info
 
 
@@ -1045,6 +1052,13 @@ def _check_helium_info(helium_info, *, info):
         ),
         "helium_info",
     )
+    if isinstance(helium_info, dict):
+        if "meas_date" in helium_info:
+            _validate_type(
+                helium_info["meas_date"],
+                datetime.datetime,
+                "helium_info['meas_date']",
+            )
     return helium_info
 
 
@@ -1331,9 +1345,13 @@ class Info(dict, SetChannelsMixin, MontageMixin, ContainsMixin):
             Helium level (%) after position correction.
         orig_file_guid : str
             Original file GUID.
-        meas_date : tuple of int
+        meas_date : datetime.datetime
             The helium level meas date.
 
+            .. versionchanged:: 1.8
+               This is stored as a :class:`~python:datetime.datetime` object
+               instead of a tuple of seconds/microseconds.
+
     * ``hpi_meas`` list of dict:
 
         creator : str
@@ -1446,8 +1464,12 @@ class Info(dict, SetChannelsMixin, MontageMixin, ContainsMixin):
             First name.
         middle_name : str
             Middle name.
-        birthday : tuple of int
-            Birthday in (year, month, day) format.
+        meas_date : datetime.date
+            The subject birthday.
+
+            .. versionchanged:: 1.8
+               This is stored as a :class:`~python:datetime.date` object
+               instead of a tuple of seconds/microseconds.
         sex : int
             Subject sex (0=unknown, 1=male, 2=female).
         hand : int
@@ -2406,7 +2428,9 @@ def read_meas_info(fid, tree, clean_bads=False, verbose=None):
                 hi["orig_file_guid"] = str(tag.data)
             elif kind == FIFF.FIFF_MEAS_DATE:
                 tag = read_tag(fid, pos)
-                hi["meas_date"] = tuple(int(t) for t in tag.data)
+                hi["meas_date"] = _ensure_meas_date_none_or_dt(
+                    tuple(int(t) for t in tag.data),
+                )
     info["helium_info"] = hi
     del hi
 
@@ -2792,7 +2816,7 @@ def write_meas_info(fid, info, data_type=None, reset_range=True):
             write_float(fid, FIFF.FIFF_HELIUM_LEVEL, hi["helium_level"])
         if hi.get("orig_file_guid") is not None:
             write_string(fid, FIFF.FIFF_ORIG_FILE_GUID, hi["orig_file_guid"])
-        write_int(fid, FIFF.FIFF_MEAS_DATE, hi["meas_date"])
+        write_int(fid, FIFF.FIFF_MEAS_DATE, _dt_to_stamp(hi["meas_date"]))
         end_block(fid, FIFF.FIFFB_HELIUM)
         del hi
 
@@ -3402,13 +3426,7 @@ def anonymize_info(info, daysback=None, keep_his=False, verbose=None):
         if none_meas_date:
             subject_info.pop("birthday", None)
         elif subject_info.get("birthday") is not None:
-            dob = datetime.datetime(
-                subject_info["birthday"][0],
-                subject_info["birthday"][1],
-                subject_info["birthday"][2],
-            )
-            dob -= delta_t
-            subject_info["birthday"] = dob.year, dob.month, dob.day
+            subject_info["birthday"] = subject_info["birthday"] - delta_t
 
         for key in ("weight", "height"):
             if subject_info.get(key) is not None:
@@ -3445,9 +3463,11 @@ def anonymize_info(info, daysback=None, keep_his=False, verbose=None):
         if hi.get("orig_file_guid") is not None:
             hi["orig_file_guid"] = default_str
         if none_meas_date and hi.get("meas_date") is not None:
-            hi["meas_date"] = DATE_NONE
+            hi["meas_date"] = _ensure_meas_date_none_or_dt(DATE_NONE)
         elif hi.get("meas_date") is not None:
-            hi["meas_date"] = _add_timedelta_to_stamp(hi["meas_date"], -delta_t)
+            hi["meas_date"] = _ensure_meas_date_none_or_dt(
+                _add_timedelta_to_stamp(hi["meas_date"], -delta_t)
+            )
 
     di = info.get("device_info")
     if di is not None:

mne/_fiff/tag.py

@@ -15,7 +15,7 @@
 from scipy.sparse import csc_array, csr_array
 
 from ..utils import _check_option, warn
-from ..utils.numerics import _julian_to_cal
+from ..utils.numerics import _julian_to_date
 from .constants import (
     FIFF,
     _ch_coil_type_named,
@@ -371,7 +371,7 @@ def _read_dir_entry_struct(fid, tag, shape, rlims):
 
 def _read_julian(fid, tag, shape, rlims):
     """Read julian tag."""
-    return _julian_to_cal(int(np.frombuffer(fid.read(4), dtype=">i4").item()))
+    return _julian_to_date(int(np.frombuffer(fid.read(4), dtype=">i4").item()))
 
 
 # Read types call dict

mne/_fiff/tests/test_meas_info.py

@@ -592,7 +592,7 @@ def _test_anonymize_info(base_info):
             his_id="foobar",
             last_name="bar",
             first_name="bar",
-            birthday=(1987, 4, 8),
+            birthday=date(1987, 4, 8),
             sex=0,
             hand=1,
         )
@@ -616,7 +616,7 @@ def _test_anonymize_info(base_info):
     # this bday is 3653 days different. the change in day is due to a
     # different number of leap days between 1987 and 1977 than between
     # 2010 and 2000.
-    exp_info["subject_info"]["birthday"] = (1977, 4, 7)
+    exp_info["subject_info"]["birthday"] = date(1977, 4, 7)
     exp_info["meas_date"] = default_anon_dos
     exp_info._unlocked = False
 
@@ -644,7 +644,7 @@ def _test_anonymize_info(base_info):
     # exp 3 tests is a supplied daysback
     delta_t_2 = timedelta(days=43)
     with exp_info_3._unlock():
-        exp_info_3["subject_info"]["birthday"] = (1987, 2, 24)
+        exp_info_3["subject_info"]["birthday"] = date(1987, 2, 24)
         exp_info_3["meas_date"] = meas_date - delta_t_2
     for key in ("file_id", "meas_id"):
         value = exp_info_3.get(key)
@@ -860,12 +860,19 @@ def test_field_round_trip(tmp_path):
             info[key] = _generate_meas_id()
         info["device_info"] = dict(type="a", model="b", serial="c", site="d")
         info["helium_info"] = dict(
-            he_level_raw=1.0, helium_level=2.0, orig_file_guid="e", meas_date=(1, 2)
+            he_level_raw=1.0,
+            helium_level=2.0,
+            orig_file_guid="e",
+            meas_date=_stamp_to_dt((1, 2)),
         )
     fname = tmp_path / "temp-info.fif"
     write_info(fname, info)
     info_read = read_info(fname)
     assert_object_equal(info, info_read)
+    info["helium_info"]["meas_date"] = (1, 2)
+    with pytest.raises(TypeError, match="datetime"):
+        # trigger the check
+        info["helium_info"] = info["helium_info"]
 
 
 def test_equalize_channels():

mne/_fiff/write.py

@@ -4,6 +4,7 @@
 # License: BSD-3-Clause
 # Copyright the MNE-Python contributors.
 
+import datetime
 import os.path as op
 import re
 import time
@@ -15,7 +16,7 @@
 from scipy.sparse import csc_array, csr_array
 
 from ..utils import _file_like, _validate_type, logger
-from ..utils.numerics import _cal_to_julian
+from ..utils.numerics import _date_to_julian
 from .constants import FIFF
 
 # We choose a "magic" date to store (because meas_date is obligatory)
@@ -120,9 +121,9 @@ def write_complex128(fid, kind, data):
 
 def write_julian(fid, kind, data):
     """Write a Julian-formatted date to a FIF file."""
-    assert len(data) == 3
+    assert isinstance(data, datetime.date), type(data)
     data_size = 4
-    jd = np.sum(_cal_to_julian(*data))
+    jd = _date_to_julian(data)
     data = np.array(jd, dtype=">i4")
     _write(fid, data, kind, data_size, FIFF.FIFFT_JULIAN, ">i4")
 

mne/export/_edf.py

@@ -164,8 +164,6 @@ def _export_raw(fname, raw, physical_range, add_ch_type):
         name = "_".join(filter(None, [first_name, middle_name, last_name]))
 
         birthday = subj_info.get("birthday")
-        if birthday is not None:
-            birthday = dt.date(*birthday)
         hand = subj_info.get("hand")
         weight = subj_info.get("weight")
         height = subj_info.get("height")

mne/export/tests/test_export.py

@@ -5,7 +5,7 @@
 # Copyright the MNE-Python contributors.
 
 from contextlib import nullcontext
-from datetime import datetime, timezone
+from datetime import date, datetime, timezone
 from pathlib import Path
 
 import numpy as np
@@ -160,7 +160,7 @@ def test_double_export_edf(tmp_path):
         his_id="12345",
         first_name="mne",
         last_name="python",
-        birthday=(1992, 1, 20),
+        birthday=date(1992, 1, 20),
         sex=1,
         weight=78.3,
         height=1.75,
@@ -290,7 +290,7 @@ def test_rawarray_edf(tmp_path):
     raw.info["subject_info"] = dict(
         first_name="mne",
         last_name="python",
-        birthday=(1992, 1, 20),
+        birthday=date(1992, 1, 20),
         sex=1,
         hand=3,
     )

mne/io/edf/edf.py

@@ -13,7 +13,7 @@
 
 import os
 import re
-from datetime import datetime, timedelta, timezone
+from datetime import date, datetime, timedelta, timezone
 
 import numpy as np
 from scipy.interpolate import interp1d
@@ -690,7 +690,7 @@ def _get_info(
             info["subject_info"]["last_name"] = sub_names[2]
     # Birthday in (year, month, day) format.
     if isinstance(edf_info["subject_info"].get("birthday"), datetime):
-        info["subject_info"]["birthday"] = (
+        info["subject_info"]["birthday"] = date(
             edf_info["subject_info"]["birthday"].year,
             edf_info["subject_info"]["birthday"].month,
             edf_info["subject_info"]["birthday"].day,
@@ -704,6 +704,9 @@ def _get_info(
     # Weight in kilograms.
     if edf_info["subject_info"].get("weight") is not None:
         info["subject_info"]["weight"] = float(edf_info["subject_info"]["weight"])
+    # Remove values after conversion to help with in-memory anonymization
+    for key in ("subject_info", "meas_date"):
+        del edf_info[key]
 
     # Filter settings
     if filt_ch_idxs := [x for x in range(len(sel)) if x not in stim_channel_idxs]:

mne/io/edf/tests/test_edf.py

@@ -129,24 +129,12 @@ def test_subject_info(tmp_path):
     want = {
         "his_id": "X",
         "sex": 1,
-        "birthday": (1967, 10, 9),
+        "birthday": datetime.date(1967, 10, 9),
         "last_name": "X",
     }
     for key, val in want.items():
         assert raw.info["subject_info"][key] == val, key
 
-    # check "subject_info" from `_raw_extras`
-    edf_info = raw._raw_extras[0]
-    assert edf_info["subject_info"] is not None
-    want = {
-        "id": "X",
-        "sex": "M",
-        "birthday": datetime.datetime(1967, 10, 9, 0, 0),
-        "name": "X",
-    }
-    for key, val in want.items():
-        assert edf_info["subject_info"][key] == val, key
-
     # add information
     raw.info["subject_info"]["hand"] = 0
 
@@ -160,7 +148,7 @@ def test_subject_info(tmp_path):
     want = {
         "his_id": "X",
         "sex": 1,
-        "birthday": (1967, 10, 9),
+        "birthday": datetime.date(1967, 10, 9),
         "last_name": "X",
         "hand": 0,
     }
@@ -1195,3 +1183,16 @@ def test_ch_types():
     raw = read_raw_edf(edf_chtypes_path, units="uV")  # should be okay
     data_units = raw.get_data()
     assert_allclose(data, data_units)
+
+
+@testing.requires_testing_data
+def test_anonymization():
+    """Test that RawEDF anonymizes data in memory."""
+    # gh-11966
+    raw = read_raw_edf(edf_stim_resamp_path)
+    for key in ("meas_date", "subject_info"):
+        assert key not in raw._raw_extras[0]
+    bday = raw.info["subject_info"]["birthday"]
+    assert bday == datetime.date(1967, 10, 9)
+    raw.anonymize()
+    assert raw.info["subject_info"]["birthday"] != bday

mne/io/edf/tests/test_gdf.py

@@ -5,7 +5,7 @@
 # Copyright the MNE-Python contributors.
 
 import shutil
-from datetime import datetime, timedelta, timezone
+from datetime import date, datetime, timedelta, timezone
 
 import numpy as np
 import scipy.io as sio
@@ -99,11 +99,10 @@ def test_gdf2_birthday(tmp_path):
         fid.seek(176, 0)
         assert np.fromfile(fid, np.uint64, 1)[0] == d
     raw = read_raw_gdf(new_fname, eog=None, misc=None, preload=True)
-    assert raw._raw_extras[0]["subject_info"]["age"] == 44
+    assert "subject_info" not in raw._raw_extras[0]
     assert raw.info["subject_info"] is not None
-
     birthdate = datetime(1, 1, 1, tzinfo=timezone.utc) + offset_44_yr
-    assert raw.info["subject_info"]["birthday"] == (
+    assert raw.info["subject_info"]["birthday"] == date(
         birthdate.year,
         birthdate.month,
         birthdate.day,
@@ -119,7 +118,7 @@ def test_gdf2_data():
         misc=None,
         preload=True,
     )
-    assert raw._raw_extras[0]["subject_info"]["age"] is None
+    assert raw.info["subject_info"]["birthday"] == date(1, 1, 1)
 
     picks = pick_types(raw.info, meg=False, eeg=True, exclude="bads")
     data, _ = raw[picks]

mne/io/fiff/tests/test_raw_fiff.py

@@ -4,6 +4,7 @@
 # License: BSD-3-Clause
 # Copyright the MNE-Python contributors.
 
+import datetime
 import os
 import pathlib
 import pickle
@@ -189,7 +190,7 @@ def test_subject_info(tmp_path):
     assert raw.info["subject_info"] is None
     # fake some subject data
     keys = ["id", "his_id", "last_name", "first_name", "birthday", "sex", "hand"]
-    vals = [1, "foobar", "bar", "foo", (1901, 2, 3), 0, 1]
+    vals = [1, "foobar", "bar", "foo", datetime.date(1901, 2, 3), 0, 1]
     subject_info = dict()
     for key, val in zip(keys, vals):
         subject_info[key] = val

mne/io/hitachi/hitachi.py

@@ -292,7 +292,7 @@ def _get_hitachi_info(fname, S_offset, D_offset, ignore_names):
     last_samp = len(bounds) - 2
 
     if age is not None and meas_date is not None:
-        subject_info["birthday"] = (
+        subject_info["birthday"] = dt.date(
             meas_date.year - age,
             meas_date.month,
             meas_date.day,

mne/io/nirx/nirx.py

@@ -343,7 +343,7 @@ def __init__(self, fname, saturated, preload=False, verbose=None):
         else:
             subject_info["sex"] = FIFF.FIFFV_SUBJ_SEX_UNKNOWN
         if inf["age"] != "":
-            subject_info["birthday"] = (
+            subject_info["birthday"] = dt.date(
                 meas_date.year - int(inf["age"]),
                 meas_date.month,
                 meas_date.day,