Cannot read some EDF files, raises ValueError: tmin should be positive

[guiweber]

I am using the latest MNE V0.19dev to read EEG data from the CHB-MIT database, located at https://alpha.physionet.org/content/chbmit/1.0.0/

Some files can be opened just fine with mne.io.read_raw_edf but other raise ValueError: tmin should be positive.

This data set has been used extensively in the literature, so I presume the files should be usable. Also, the traceback shows that the error originates in crop, called by set_annotations, which indicates that the problematic values are in annotations, which shouldn't prevent using the data.

Steps to reproduce

1. Download https://alpha.physionet.org/static/published-projects/chbmit/1.0.0/chb06/chb06_03.edf
2. Try opening the file with mne.io.read_raw_edf

Expected results

If the file parameters or annotations are wrong but the data is otherwise ok, the parameters/annotations should be set to reasonable default values and a warning should be produced. It should still be possible to load the file.

Actual results

It is impossible to load the file

Debug output and traceback

raw = mne.io.read_raw_edf(file, preload=True, verbose='DEBUG')

Extracting EDF parameters from xxx\chb06_03.edf... EDF file detected xxx/main.py:19: RuntimeWarning: Channel names are not unique, found duplicates for: {'T8-P8'}. Applying running numbers for duplicates. Setting channel info structure... FP1-F7: range=1600.0 cal=4095.0 F7-T7: range=1600.0 cal=4095.0 T7-P7: range=1600.0 cal=4095.0 P7-O1: range=1600.0 cal=4095.0 FP1-F3: range=1600.0 cal=4095.0 F3-C3: range=1600.0 cal=4095.0 C3-P3: range=1600.0 cal=4095.0 P3-O1: range=1600.0 cal=4095.0 FP2-F4: range=1600.0 cal=4095.0 F4-C4: range=1600.0 cal=4095.0 C4-P4: range=1600.0 cal=4095.0 P4-O2: range=1600.0 cal=4095.0 FP2-F8: range=1600.0 cal=4095.0 F8-T8: range=1600.0 cal=4095.0 T8-P8-0: range=1600.0 cal=4095.0 P8-O2: range=1600.0 cal=4095.0 FZ-CZ: range=1600.0 cal=4095.0 CZ-PZ: range=1600.0 cal=4095.0 P7-T7: range=1600.0 cal=4095.0 T7-FT9: range=1600.0 cal=4095.0 FT9-FT10: range=1600.0 cal=4095.0 FT10-T8: range=1600.0 cal=4095.0 T8-P8-1: range=1600.0 cal=4095.0 Creating raw.info structure... Reading 0 ... 3686399 = 0.000 ... 14399.996 secs... Traceback (most recent call last): File "", line 1, in File "C:\Program Files\JetBrains\PyCharm Community Edition 2019.1.2\helpers\pydev_pydev_bundle\pydev_umd.py", line 197, in runfile pydev_imports.execfile(filename, global_vars, local_vars) # execute the script File "C:\Program Files\JetBrains\PyCharm Community Edition 2019.1.2\helpers\pydev_pydev_imps_pydev_execfile.py", line 18, in execfile exec(compile(contents+"\n", file, 'exec'), glob, loc) File "xxx/main.py", line 19, in raw = mne.io.read_raw_edf(file, preload=True, verbose='DEBUG') File "C:\Program Files\Python37\lib\site-packages\mne\io\edf\edf.py", line 1244, in read_raw_edf verbose=verbose) File "<C:\Program Files\Python37\lib\site-packages\mne\externals\decorator.py:decorator-gen-152>", line 2, in init File "C:\Program Files\Python37\lib\site-packages\mne\utils_logging.py", line 88, in wrapper return function(*args, **kwargs) File "C:\Program Files\Python37\lib\site-packages\mne\io\edf\edf.py", line 193, in init description=desc, orig_time=None)) File "C:\Program Files\Python37\lib\site-packages\mne\io\base.py", line 795, in set_annotations emit_warning=emit_warning) File "C:\Program Files\Python37\lib\site-packages\mne\annotations.py", line 361, in crop raise ValueError('tmin should be positive.') ValueError: tmin should be positive.

Additional information

Platform: Windows-10-10.0.17763-SP0 Python: 3.7.2 (tags/v3.7.2:9a3ffc0492, Dec 23 2018, 23:09:28) [MSC v.1916 64 bit (AMD64)] Executable: xxx\venv\Scripts\python.exe CPU: Intel64 Family 6 Model 94 Stepping 3, GenuineIntel: 8 cores Memory: Unavailable (requires "psutil" package) mne: 0.19.dev0 numpy: 1.16.1 {blas=C:\projects\numpy-wheels\numpy\build\openblas, lapack=C:\projects\numpy-wheels\numpy\build\openblas} scipy: 1.2.1 matplotlib: 3.0.3 {backend=TkAgg} sklearn: 0.21.1 nibabel: Not found mayavi: Not found cupy: Not found pandas: 0.24.2 dipy: Not found

[larsoner]

@massich can you look?

[cbrnr]

The file does not have any annotations. The problem is likely the start of recording date, which is set to 2060-01-27 03:09:42.

[cbrnr]

Or rather, 1960 instead of 2060 (2060 is returned by biosig, 1960 by Python or R datetime functions).

[agramfort]

I confirm that #6391 fixes the bug.

shall we merge and backport to 0.18 branch?

[massich]

Yes, #6391 fixes the bug, and I guess that we could backport it. I'll do it today

[Matteo73]

@massich Thanks for backporting it to 0.18. When the 0.18 updated version will be avaible?

Cheers, Matteo

[agramfort]

can you rely on master for your work?

[Matteo73]

@agramfort Yes, of course, I can. I will migrate to the stable version when it is available.

Cheers, Matteo