Datetime on the header file is restricted with DD/MM/YYYY format

[zeynep-ozturk]

Hi,
I wanted to use your package to make peak detection and remove them. However, in my case the header files are like this:

A00001 1 300 9000 2014-05-08 05:05:15
A00001.mat 16+24 1000/mV 16 0 -127 0 0 ECG

This is from the 2017 challenge by the way. Because your code only supports DD/MM/YYYY format, I am getting this error:

File "/Users/zeynepozturk/dersler/ie517_hw3/venv/lib/python3.6/site-packages/wfdb/io/record.py", line 1130, in rdrecord
record = rdheader(record_name, pb_dir=pb_dir, rd_segments=False)
File "/Users/zeynepozturk/dersler/ie517_hw3/venv/lib/python3.6/site-packages/wfdb/io/record.py", line 981, in rdheader
record_fields = _header._read_record_line(header_lines[0])
File "/Users/zeynepozturk/dersler/ie517_hw3/venv/lib/python3.6/site-packages/wfdb/io/_header.py", line 743, in _read_record_line
record_fields['base_time'] = wfdb_strptime(record_fields['base_time'])
File "/Users/zeynepozturk/dersler/ie517_hw3/venv/lib/python3.6/site-packages/wfdb/io/_header.py", line 672, in wfdb_strptime
return datetime.datetime.strptime(time_string, time_fmt).time()
File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/_strptime.py", line 565, in _strptime_datetime
tt, fraction = _strptime(data_string, format)
File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/_strptime.py", line 365, in _strptime
data_string[found.end():])
ValueError: unconverted data remains: 14
I am using version '2.0.4'.

This is my first issue on Github, so I apologize beforehand for my mistakes.

Hope you can improve this and be open to new date formats.

Best regards,

Zeynep

[cx1111]

The specifications are stated here, and I just follow them: https://www.physionet.org/physiotools/wag/header-5.htm

Those files are technically wrong according to the standards, and I won't change the parser to allow for erroneous files. I'll change the files and let you know.

[zeynep-ozturk]

The files were from PhysioNet/CinC Challenge 2017. Thanks. https://physionet.org/physiobank/database/challenge/2017/

[efemaer]

I found where the problem is, the format in the .hea files is wrong - maybe outdated?

Here is the proper format for record A00001:

A00001 1 300 9000 05:05:15 8/5/2014
A00001.mat 16+24 1000/mV 16 0 -127 0 0 ECG

I will try to figure a script to edit all those files and fix them.

[efemaer]

This did the trick for me:

#!/usr/bin/python3

import glob, os
import shutil
import re

os.chdir('PATH-TO-YOUR-DATASET/')

for file in glob.glob('*.hea'):
    with open(file) as from_file:

        line = from_file.readline().strip()

        # make any changes to line here
        p = re.compile('(?P<name>\S+)\s(?P<num>\S+)\s(?P<freq>\S+)\s(?P<length>\S+)\s(?P<year>\d+)-(?P<month>\d+)-(?P<day>\d+)\s(?P<hr>\d+):(?P<min>\d+):(?P<sec>\d+)')
        info = p.search(line)

        new_line = '{} {} {} {} {}:{}:{} {}/{}/{} \n'.format(info.group('name'), info.group('num'), info.group('freq'), info.group('length'), info.group('hr'), info.group('min'), info.group('sec'), info.group('day'), info.group('month'), info.group('year'))

        with open(file, 'w') as to_file:
            to_file.write(new_line)
            shutil.copyfileobj(from_file, to_file)

PS: it takes a bit of time, since it has to open each file, edit, then copy the rest of the file

EDIT: A05967.hea for some reason is the only file missing the time, so just add 00:00:00 at the end of the first line before running the script

[cx1111]

My coworker has changed the dates. Thanks.