[MRG] Egimff by ramonapariciog · Pull Request #4017 · mne-tools/mne-python

ramonapariciog · 2017-02-24T19:51:44Z

mne-mff-reader BETA

MNE Class and functions for loading the EGI '.mff' files from EGI Netstation EEG.

How to import:
Only with the following line:
from mne.io.mff import read_raw_egi_mff

and for create the raw instance:
raw = read_raw_egi_mff(filepath, exclude, include, verbose)

All the pararameters are the same than for the read_raw_egi.
Limitations:

1.- At this moment doesn't support multisubject records, but can be adapted for that.

2.- At the moment has been tested by comparing the results with the obtained by a single subject and notsegmented record raw file, and the tests in the test_egi.py

3.- At the moment all data is loaded to memory, using the preload=True mode.

Thanks to PhD. Guillaume Dumas, PhD. Dennis A. Engeman & PhD. Sheraz Khan

larsoner · 2017-02-24T19:53:21Z

Those limitations all seem reasonable, we can expand later once this is in.

Is this ready for a round of review?

agramfort

there is much more but it's a start

agramfort · 2017-02-26T15:23:00Z

mne/io/egi/egimff.py

+        Path for the file
+    Returns
+    ------------------------
+    info : dictionary


dictionary -> dict

agramfort · 2017-02-26T15:23:22Z

mne/io/egi/egimff.py

+    input_fname : str
+        Path for the file
+    Returns
+    ------------------------


Returns ------------------------ -> Returns --------

agramfort · 2017-02-26T15:23:51Z

mne/io/egi/egimff.py

+    #    version = version.byteswap().astype(np.uint32)
+    # else:
+    #    ValueError('Watchout. This does not seem to be a simple '
+    #               'binary EGI file.')


to cleanup?

agramfort · 2017-02-26T15:24:16Z

mne/io/egi/egimff.py

+def _read_events(input_fname, hdr, info):
+    """Read events for the record.
+
+    in:


to format properly

agramfort · 2017-02-26T15:25:00Z

mne/io/egi/egimff.py

+
+
+@verbose
+def read_raw_egi_mff(input_fname, montage=None, eog=None, misc=None,


why not factorize this with the read_raw_egi to based on extension decide if you use mff or not?

agramfort · 2017-02-26T15:28:54Z

mne/io/egi/source/general.py

+def _get_signalfname(filepath, infontype):
+    import os
+    from xml.dom.minidom import parse
+    import re


to import at the top of the file

agramfort · 2017-02-26T15:29:00Z

mne/io/egi/source/general.py

+##
+
+def _u2sample(microsecs, samprate):
+    import numpy as np


agramfort · 2017-02-26T15:29:13Z

mne/io/egi/source/general.py

+#  sampDuration = 1000000/sampRate;
+#  sampleNum = microsecs/sampDuration;
+#  remainder = uint64(rem(microsecs, sampDuration));
+#  sampleNum = fix(sampleNum);


agramfort · 2017-02-26T15:29:23Z

mne/io/egi/source/general.py

+    """Read data signal."""
+    import numpy as np
+
+    binfile = filepath + '/' + nbinfile


os.path.join

agramfort · 2017-02-26T15:30:24Z

mne/io/egi/source/header.py

+        blockbeginsamps[x + 1] = blockbeginsamps[x] + blocknumsamps[x]
+    #  -----------------------------------------------------
+    summaryinfo = dict(blocks=signal_blocks['blocks'],
+                       eegFilename=signal_blocks['eegFile'],


if don't use CamelCase for variable names but snake_case

agramfort · 2017-02-26T15:32:38Z

@ramonapariciog thanks for taking a stab at this. There is a bit of work
to complete this. Maybe an easy way would be to come for a pair programing day with
@jaeilepp at Télécom?

SherazKhan · 2017-03-02T19:05:28Z

@ramonapariciog excellent Job, @agramfort Ramon did great Job, all camelCases was my bad :-)

SherazKhan · 2017-03-02T19:07:11Z

@ramonapariciog @Eric89GXL @agramfort may be we all work on this during hackathon ?

dengemann · 2017-03-02T19:12:32Z

Sheraz, I think Roman and Jaakko will meet soon to finalize this one in a pair programming session :) So it will be merged rather soon I predict.

…

On Thu, Mar 2, 2017 at 8:07 PM Sheraz Khan ***@***.***> wrote: @ramonapariciog <https://github.com/ramonapariciog> @Eric89GXL <https://github.com/Eric89GXL> @agramfort <https://github.com/agramfort> may be we all work on this during hackathon ? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#4017 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AB0finMZAiWPW_cN7Zczkcpk44_oLgzhks5rhxNggaJpZM4MLm6H> .

SherazKhan · 2017-03-02T20:12:50Z

@dengemann you guys are awesome :)

jaeilepp · 2017-03-15T13:59:27Z

Now the preload=False case works too. I don't dare to remove the epoch reading part although I think it should be made it's own function (read_epochs_egimff or something). I'll leave that to @ramonapariciog since I don't have the data to test it.

larsoner

I'll stop reviewing for now. I pointed out a few small pattern things, but I suspect you were already planning to clean up most of them.

larsoner · 2017-03-15T13:58:54Z

mne/io/egi/egimff.py

+"""
+Created on Fri Jan 13 11:01:39 2017.
+
+@author: ramonapariciog


clean up docstring

larsoner · 2017-03-15T13:59:24Z

mne/io/egi/egimff.py

+    time_n = dateutil.parser.parse(mff_hdr['date'])
+    info = dict(   # leer del info.xml desde read_mff_header
+        version=version,  # duda
+        year=int(time_n.strftime('%Y')),           # duda


That's the current president of Poland, maybe related? 😉

larsoner · 2017-03-15T13:59:47Z

mne/io/egi/egimff.py

+
+
+@verbose
+def read_raw_egi_mff(input_fname, montage=None, eog=None, misc=None,


larsoner · 2017-03-15T14:01:22Z

mne/io/egi/egimff.py

+              on timestamps received by the Netstation. As a consequence,
+              triggers have only short durations.
+
+              This step will fail if events are not mutually exclusive.


This .. note:: is very long. Let's move it to the Notes section. If you want, you can add a on or two sentence .. note:: here like "this function will attempt to generate a synthetic trigger channel. For details, see Notes below.".

larsoner · 2017-03-15T14:01:43Z

mne/io/egi/egimff.py

+    montage : str | None | instance of montage
+        Path or instance of montage containing electrode positions.
+        If None, sensor locations are (0,0,0). See the documentation of
+        :func:`mne.channels.read_montage` for more information.


I think we should stop putting these in the constructor and have people use the methods instead

hasn't been addressed yet

Now this is fused with the egi reader. Deprecation should be done in another PR

larsoner · 2017-03-15T14:02:46Z

mne/io/egi/egimff.py

+            if len(data) != len(info['ch_names']):
+                raise ValueError('len(data) does not match '
+                                 'len(info["ch_names"])')
+            logger.info('Creating RawArray with %s data, n_channels=%s, '


This isn't true anymore?

larsoner · 2017-03-15T14:03:14Z

mne/io/egi/source/data.py

+
+
+def read_mff_data(filepath, indtype, startind, lastind, hdr):
+    """Function for load the data in a list.


Don't say "function for" or "helper" anymore, use imperative

larsoner · 2017-03-15T14:03:50Z

mne/io/egi/source/data.py

+    """
+    import numpy as np
+    import os
+    from .general import _bls2blns, _read_signaln, _get_gains


don't nest these

hopefully general does not create a circular import. If it does, some helper function probably needs to be moved

larsoner · 2017-03-15T14:03:58Z

mne/io/egi/source/data.py

+        info_fp = os.path.join(filepath, suminfo['infoFile'][0])
+        gains = _get_gains(info_fp)
+    else:
+        print('Multisubject not suported')


don't print

jaeilepp · 2017-04-18T14:35:19Z

I can take a look tomorrow. So the test_egi.mff is supposedly continuous data, not epoched?

ramonapariciog · 2017-04-18T14:46:55Z

Hello Jakko,
Yes, it was took as continuous data.

ramonapariciog · 2017-04-18T14:52:17Z

Thank you Jakko, if you need something i will try to respond you quickly.

jaeilepp · 2017-04-19T09:40:39Z

Now it should work. I'm a bit worried about the event timing. It seems that the event times in the test data are all smaller than the meas_date. I have no idea how to sync them. It also seems that the it doesn't give any meaningful evoked responses when using the current event timing.

jaeilepp · 2017-04-19T09:41:30Z

mne/io/egi/egimff.py

+                warn('Event outside data range (%ss).' % (i /
+                                                          info['samp_rate']))
+                continue
+            events[n][i] = 2**n


I changed it to use powers of 2 so it'll be easy to mask overlaps.

Does this create inconsistency with how the other readers work, or do we not assign consecutive values elsewhere?

Yeah, I noticed that before it was doing basically the same thing as the egi reader (consecutive values). The problem is that it doesn't allow overlapping events.

Okay then we should probably have a parameter for this. event_numbering='linear' | 'log' or 'consecutive' | 'power' or something. If we want to make 2**n the default here, it should probably be part of this PR, or as part of an immediate follow-up.

jaeilepp · 2017-04-19T09:42:31Z

mne/io/egi/egimff.py

+        # STI 014 is simply the sum of all event channels (powers of 2).
+        if len(egi_events) > 0:
+            egi_events = np.vstack([egi_events, np.sum(egi_events, axis=0)])
+            data[n_channels:] = egi_events


I also made the stim channel construction lazy like this. Now it's created on data fetch stage.

Ok Jakko, thanks i will test it with other files.

jaeilepp · 2017-04-19T09:43:02Z

ping @ramonapariciog

jaeilepp · 2017-04-20T06:59:10Z

mne/io/egi/source/general.py

+            nsamples = block['nsamples']
+        else:
+            raise NotImplementedError('Only continuos files are supported')
+        numblocks = int(l // blocksize)


@ramonapariciog the problem was here. It was previously not doing integer division and then adding an extra block if the data did not align. If I understand correctly, the proper thing to do is to just dump the extra data, since it is only available for the first channels in the block.

jona-sassenhagen · 2017-04-20T22:07:42Z

I'll just say I'd appreciate no longer having to use EEGLAB to import MFF files :)

jaeilepp · 2017-04-24T13:03:30Z

There's still a lot of cleaning to do. Seems like there's a lot of redundant meta data stored. I think the best is to store only the necessary data and then add more later if needed.

dengemann · 2017-04-24T13:25:55Z

+1 be pragmatic. Store what is needed for conversion.

…

On Mon, 24 Apr 2017 at 09:03, jaeilepp ***@***.***> wrote: There's still a lot of cleaning to do. Seems like there's a lot of redundant meta data stored. I think the best is to store only the necessary data and then add more later if needed. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#4017 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AB0fihEaEgIh16vmcE81ygNwrch7pGU1ks5rzJ2igaJpZM4MLm6H> .

ramonapariciog · 2017-04-24T14:52:22Z

Sure, i will check which data is needed from the first version to discart it.

ramonapariciog · 2017-04-26T19:02:27Z

Hi Jakko, i have problems to use the function again :'(

mne.io.read_raw_egi_mff('mmnbenjamin23022017_20170223_111927.mff', preload=True)
Reading EGI MFF Header from /Users/ghfc/Desktop/mmn/mmnbenjamin23022017_20170223_111927.mff...
    Reading events ...
    Assembling measurement info ...
    Synthesizing trigger channel "STI 014" ...
    Excluding events {} ...
Reading 0 ... 874358  =      0.000 ...   874.358 secs...
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-3-a83585f0f406> in <module>()
----> 1 mne.io.read_raw_egi_mff('/Users/ghfc/Desktop/mmn/mmnbenjamin23022017_20170223_111927.mff', preload=True)

/Users/ghfc/mne-python/mne/io/egi/egimff.py in read_raw_egi_mff(input_fname, montage, eog, misc, include, exclude, preload, verbose)

/Users/ghfc/mne-python/mne/utils.py in verbose(function, *args, **kwargs)
    705         with use_log_level(verbose_level):
    706             return function(*args, **kwargs)
--> 707     return function(*args, **kwargs)
    708
    709

/Users/ghfc/mne-python/mne/io/egi/egimff.py in read_raw_egi_mff(input_fname, montage, eog, misc, include, exclude, preload, verbose)
    234     """
    235     return RawMff(input_fname, montage, eog, misc, include, exclude,
--> 236                   preload, verbose)
    237
    238

/Users/ghfc/mne-python/mne/io/egi/egimff.py in __init__(self, input_fname, montage, eog, misc, include, exclude, preload, verbose)

/Users/ghfc/mne-python/mne/utils.py in verbose(function, *args, **kwargs)
    705         with use_log_level(verbose_level):
    706             return function(*args, **kwargs)
--> 707     return function(*args, **kwargs)
    708
    709

/Users/ghfc/mne-python/mne/io/egi/egimff.py in __init__(self, input_fname, montage, eog, misc, include, exclude, preload, verbose)
    339             info, preload=preload, orig_format=egi_info['orig_format'],
    340             filenames=[file_bin], last_samps=[egi_info['n_samples'] - 1],
--> 341             raw_extras=[egi_info], verbose=verbose)
    342
    343     def _read_segment_file(self, data, idx, fi, start, stop, cals, mult):

/Users/ghfc/mne-python/mne/io/base.py in __init__(self, info, preload, first_samps, last_samps, filenames, raw_extras, orig_format, dtype, verbose)

/Users/ghfc/mne-python/mne/utils.py in verbose(function, *args, **kwargs)
    705         with use_log_level(verbose_level):
    706             return function(*args, **kwargs)
--> 707     return function(*args, **kwargs)
    708
    709

/Users/ghfc/mne-python/mne/io/base.py in __init__(self, info, preload, first_samps, last_samps, filenames, raw_extras, orig_format, dtype, verbose)
    363         self._update_times()
    364         if load_from_disk:
--> 365             self._preload_data(preload)
    366
    367     @verbose

/Users/ghfc/mne-python/mne/io/base.py in _preload_data(self, preload, verbose)

/Users/ghfc/mne-python/mne/utils.py in verbose(function, *args, **kwargs)
    705         with use_log_level(verbose_level):
    706             return function(*args, **kwargs)
--> 707     return function(*args, **kwargs)
    708
    709

/Users/ghfc/mne-python/mne/io/base.py in _preload_data(self, preload, verbose)
    618         logger.info('Reading %d ... %d  =  %9.3f ... %9.3f secs...' %
    619                     (0, len(self.times) - 1, 0., self.times[-1]))
--> 620         self._data = self._read_segment(data_buffer=data_buffer)
    621         assert len(self._data) == self.info['nchan']
    622         self.preload = True

/Users/ghfc/mne-python/mne/io/base.py in _read_segment(self, start, stop, sel, data_buffer, projector, verbose)
    515             self._read_segment_file(data[:, this_sl], idx, fi,
    516                                     int(start_file), int(stop_file),
--> 517                                     cals, mult)
    518             offset += n_read
    519         return data

/Users/ghfc/mne-python/mne/io/egi/egimff.py in _read_segment_file(self, data, idx, fi, start, stop, cals, mult)
    383                 end = count // n_channels + 2
    384                 if sample_start == 0:
--> 385                     block = block[:, s_offset:end]
    386                     sample_sl = slice(sample_start,
    387                                       sample_start + block.shape[1])

TypeError: slice indices must be integers or None or have an __index__ method

i update numpy scipy and also, to use the new version of autoreject and this happens when i try load a file.
The versions:
numpy 1.12.1
scipy 0.19.0

jaeilepp · 2017-04-27T06:53:21Z

Strange that I can't reproduce this even after creating an environment with the same settings as travis.

codecov-io · 2017-04-27T08:32:40Z

Codecov Report

Merging #4017 into master will increase coverage by 0.03%.
The diff coverage is 90.29%.

@@            Coverage Diff             @@
##           master    #4017      +/-   ##
==========================================
+ Coverage   86.18%   86.21%   +0.03%     
==========================================
  Files         358      349       -9     
  Lines       65020    65084      +64     
  Branches     9914     9983      +69     
==========================================
+ Hits        56040    56115      +75     
+ Misses       6247     6222      -25     
- Partials     2733     2747      +14

jaeilepp · 2017-04-27T10:23:58Z

Currently this data is not used. I assume at least GCAL should be used for calibration, but I have no idea what ICAL is. EEGLAB reader seems to recognize calibrated_gains and calibrated_zeros. Could ICAL be zero calibration?

larsoner · 2017-04-27T11:08:09Z

Gain and intercept (multiplicative and additive) would make sense for GCAL and ICAL

ramonapariciog · 2017-04-27T13:34:24Z

Hello Jakko, i check again and now i think that is loading the data, but, you already tried using the data with Autoreject?
i get this error:

In [20]: epochs_clean = ar.fit(epochs[::50]).transform(epochs)
Loading data for 18 events and 701 original time points ...
1 bad epochs dropped
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
<ipython-input-20-290c8fbf6820> in <module>()
----> 1 epochs_clean = ar.fit(epochs[::50]).transform(epochs)

C:\Program Files\Anaconda3\envs\mne_dev\lib\site-packages\autoreject-0.1.dev0-py3.5.egg\autoreject\autoreject.py in fit(self, epochs)
    515             The epochs object to be fit.
    516         """
--> 517         _check_data(epochs)
    518         if self.cv is None:
    519             self.cv = KFold(len(epochs), n_folds=10)

C:\Program Files\Anaconda3\envs\mne_dev\lib\site-packages\autoreject-0.1.dev0-py3.5.egg\autoreject\autoreject.py in _check_data(epochs)
     38                'incomplete data). Please check that no epoch '
     39                'is dropped when you call epochs.drop_bad_epochs().')
---> 40         raise RuntimeError(msg)
     41
     42

RuntimeError: Some epochs are being dropped (maybe due to incomplete data). Please check that no epoch is dropped when you call epochs.drop_bad_epochs().

Do you think that could be the timeline?

jaeilepp · 2017-05-04T11:29:53Z

I think this is close to being ready. I couldn't figure out how to use the ICAL values, but the GCALs are now in use. I compared to the converted egi files I have and the data scales are very close to each other.

larsoner · 2017-05-04T15:09:24Z

Can you update manual/io.rst? It would answer the question I was about to ask, which is what's the difference between egi and egi_mff

larsoner · 2017-05-04T15:11:37Z

Regarding ICAL/GCAL can you cross-check with FieldTrip (e.g., http://www.fieldtriptoolbox.org/faq/how_can_i_read_egi_mff_data_without_the_jvm)

larsoner

Otherwise LGTM

larsoner · 2017-05-04T15:18:01Z

mne/io/egi/egimff.py

+    signal_blocks = _get_blocks(fname)
+    blocknumsamps = np.sum(signal_blocks['blockNumSamps'])
+
+    pibhasref = False


either make the name more explicit or add a comment, I don't know what pib means

larsoner · 2017-05-04T15:18:13Z

mne/io/egi/egimff.py

+
+    summaryinfo = dict(eeg_fname=eeg_file,
+                       info_fname=info_files[0],
+                       pibNChans=pibnchans,


no camelCase please

larsoner · 2017-05-04T15:18:31Z

mne/io/egi/egimff.py

+        ntrials = 1
+
+    # Add the sensor info.
+    sensor_layout_file = filepath + '/sensorLayout.xml'


larsoner · 2017-05-04T15:18:58Z

mne/io/egi/egimff.py

+    for sensor in sensors:
+        sensortype = int(sensor.getElementsByTagName('type')[0]
+                         .firstChild.data)
+        if sensortype == 0 or sensortype == 1:


can just say in [0, 1]

larsoner · 2017-05-04T15:19:40Z

mne/io/egi/egimff.py

+            if sensor.getElementsByTagName('name')[0].firstChild is None:
+                sn = sensor.getElementsByTagName('number')[0].firstChild.data
+                sn = sn.encode()
+                tmp_label = 'E' + sn.decode()


why encode then decode? why not just u'E' + text_type(sn)?

larsoner · 2017-05-04T15:21:24Z

mne/io/egi/egimff.py

+        gain=0,
+        bits=0,
+        value_range=0)
+    unsegmented = 1 if mff_hdr['nTrials'] == 1 else 0


= int(mff_hdr['nTrials'] == 1)

larsoner · 2017-05-04T15:21:38Z

mne/io/egi/egimff.py

+    unsegmented = 1 if mff_hdr['nTrials'] == 1 else 0
+    precision = 4
+    if precision == 0:
+        RuntimeError('Floating point precision is undefined.')


unsupported

... also this can never happen because precision = 4 immediately above...?

larsoner · 2017-05-04T15:22:46Z

mne/io/egi/egimff.py

+
+
+class RawMff(BaseRaw):
+    """RAWMff class."""


larsoner · 2017-05-04T15:23:25Z

mne/io/egi/egimff.py

+    def _read_segment_file(self, data, idx, fi, start, stop, cals, mult):
+        """Read a chunk of data."""
+        from ..utils import _mult_cal_one
+        dtype = '<f4'


it seems like this should be read from the header somewhere but it's not currently (and the check above is broken)

I can't find it anywhere in the meta data and it seems fieldtrip has the size hard coded too.

Sounds reasonable, would be good to add a comment along these lines

larsoner · 2017-05-04T15:25:07Z

mne/io/egi/events.py

+from glob import glob
+from os.path import basename, join, splitext
+
+from xml.etree.ElementTree import parse


this should be grouped with the standard imports

jaeilepp · 2017-05-05T06:00:56Z

I couldn't find anything about the calibrations in fieldtrip code base.

larsoner

One last structural thing. @agramfort a while ago mentioned rolling this into read_raw_egi and triaging based on file extension. Is that possible?

larsoner · 2017-05-05T13:48:54Z

mne/io/egi/egimff.py

+    montage : str | None | instance of montage
+        Path or instance of montage containing electrode positions.
+        If None, sensor locations are (0,0,0). See the documentation of
+        :func:`mne.channels.read_montage` for more information.


hasn't been addressed yet

larsoner · 2017-05-05T13:49:16Z

mne/io/egi/egimff.py

+    def _read_segment_file(self, data, idx, fi, start, stop, cals, mult):
+        """Read a chunk of data."""
+        from ..utils import _mult_cal_one
+        dtype = '<f4'


Sounds reasonable, would be good to add a comment along these lines

larsoner · 2017-05-05T13:49:48Z

mne/io/egi/events.py

+
+
+def _ns(s):
+    """Remove namespace, but only it there is a namespace to begin with."""


larsoner · 2017-05-05T13:50:13Z

mne/io/egi/tests/test_egi.py

+    assert_true('RawMff' in repr(raw))
+    include = ['DIN1', 'DIN2', 'DIN3', 'DIN4',
+               'DIN5', 'DIN7']
+    with warnings.catch_warnings(record=True):


add comment about what warnings are being caught

ramonapariciog · 2017-05-05T17:01:54Z

Hello guys,
some commentaries:
I need to know for each file the id for the events. The way as i did at the begining was to create the event array with n rows x s columns, n= number of the stimulus channels, s = number of samples. The way to order the stimulus channels is according to the order of those appears on time, and the program creates the event_id dictionary with the relation between the number of the row for each channel, and the STI 014 channel was just the sum. I think that now the function just take the sum of all the channels to create the STI 014 and the include and exclude criteria are not working. I will try to adjust this on the weekend because i need it, and i have the function that does that. During the process was deleted and we wasn't aware about its importance.

agramfort · 2017-05-23T19:29:17Z

@ramonapariciog @SherazKhan can you guys test and report any difficulty?

thanks heaps @jaeilepp for the hard work !

ramonapariciog · 2017-05-24T20:44:33Z

Yeah sure, thanks @jaeilepp for all, i will test.

agramfort · 2017-06-25T22:05:16Z

ping @ramonapariciog @SherazKhan

can you guys test and report any difficulty?

agramfort · 2017-07-11T18:39:25Z

ok let's say it's working :)

@jaeilepp please update what's new and merge

agramfort · 2017-07-11T21:14:22Z

Ok I'll update what's new in master. I feel like closing PRs...

agramfort · 2017-07-11T21:21:49Z

see 82306db

ramonapariciog · 2017-07-12T17:47:14Z

ohhh, i have problems with my computer for now, but yeah, the last version of the function was working without problems.
Sorry, @agramfort, it is good idea to delete the branch now?

agramfort · 2017-07-12T18:43:02Z

no need to delete. please report problem with what I merged in master yesterday

* Create read_egi_mff I need to understand from where part of the _BaseRaw class __init__() the _read_segment_file are called to execution, because the data from Sheraz reads the data and only adapting it is enough for a first test. * The egimff.py was added with the read_raw_egi_mff function. To import use from mne.io import read_raw_egi_mff * Correction of some PEP8 errors and delete the folder mff bad added * Raw reader. * preload=False * Cleaning. * Fixes to event reading. * Fixes. * Fixes. * cleaning useless functions * Cleaning. * Reading of meta data at the end. * Update tests. Cleaning. * Cleaning. Correct number of samples. * More cleaning. * Cleaning. * Update testing dataset. * Update test. * Fix. * Test fix. * Fixes. * Docs. Switched back to running number events. * Gains. Fix to events. * Cleaning. Apply gains. * Fixes. * Fixes. * Read coordinates. Cleaning. * Using _combine_triggers to create STI 014 and event_id attribute * Refactoring. * Fix.

agramfort reviewed Feb 26, 2017

View reviewed changes

agramfort changed the title ~~Egimff~~ [WIP] Egimff Feb 26, 2017

larsoner reviewed Mar 15, 2017

View reviewed changes

jaeilepp reviewed Apr 19, 2017

View reviewed changes

jaeilepp force-pushed the egimff branch from e31f3a5 to 6b0f4c5 Compare April 20, 2017 06:55

jaeilepp reviewed Apr 20, 2017

View reviewed changes

jaeilepp force-pushed the egimff branch from f013b52 to bb09dcb Compare April 25, 2017 13:06

jaeilepp added 3 commits May 4, 2017 08:44

Gains. Fix to events.

d88f9d0

Cleaning. Apply gains.

57bbb3c

Fixes.

47e9420

Fixes.

d2e9ae6

jaeilepp force-pushed the egimff branch from fbe44fe to d2e9ae6 Compare May 4, 2017 12:40

larsoner reviewed May 4, 2017

View reviewed changes

Read coordinates. Cleaning.

c0caa38

larsoner reviewed May 5, 2017

View reviewed changes

ramonapariciog and others added 3 commits May 6, 2017 02:39

Using _combine_triggers to create STI 014 and event_id attribute

48f9275

Refactoring.

c849f10

Fix.

74063a0

jaeilepp changed the title ~~[WIP] Egimff~~ [MRG] Egimff May 23, 2017

agramfort merged commit 7cbd69c into mne-tools:master Jul 11, 2017

mscheltienne mentioned this pull request Dec 29, 2022

Multiple issues while reading and resampling raw mff #11380

Open

drammock mentioned this pull request Jul 11, 2023

fixed excluding an event that happens first and once. #11674

Closed

larsoner mentioned this pull request Dec 4, 2023

API: Use annotations for read_raw_egi events rather than synthetic channel #12262

Closed



		@verbose
		def read_raw_egi_mff(input_fname, montage=None, eog=None, misc=None,



		def read_mff_data(filepath, indtype, startind, lastind, hdr):
		"""Function for load the data in a list.



		def _ns(s):
		"""Remove namespace, but only it there is a namespace to begin with."""

Uh oh!

Conversation

ramonapariciog commented Feb 24, 2017

Uh oh!

larsoner commented Feb 24, 2017

Uh oh!

agramfort left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

agramfort commented Feb 26, 2017

Uh oh!

SherazKhan commented Mar 2, 2017

Uh oh!

SherazKhan commented Mar 2, 2017

Uh oh!

dengemann commented Mar 2, 2017 via email

Uh oh!

SherazKhan commented Mar 2, 2017

Uh oh!

jaeilepp commented Mar 15, 2017

Uh oh!

larsoner left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jaeilepp commented Apr 18, 2017

Uh oh!

ramonapariciog commented Apr 18, 2017

Uh oh!

ramonapariciog commented Apr 18, 2017

Uh oh!

jaeilepp commented Apr 19, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ramonapariciog commented Apr 26, 2017 •

edited

Loading

codecov-io commented Apr 27, 2017 •

edited

Loading