[Mne_analysis] Fwd: epochs baseline problem (pandas dataframe indexing?)

Nakagawa Tristan T nakagawa-t at ifrec.osaka-u.ac.jp
Fri Apr 15 00:43:13 EDT 2016
Search archives:

Ah, yes, also the typo, thanks (see my other post)!


On 2016年04月14日 15:27, Jaakko Leppäkangas wrote:
> Hi Tristan,
> seems like a typo. You have baseline defined as `(-1, -.05)`, but in
> the function call you use `baseline=(-.5,.05)`. Just change that to
> `baseline=baseline`.
>
>
> On 14 April 2016 at 03:27, Nakagawa Tristan T
> <nakagawa-t at ifrec.osaka-u.ac.jp
> <mailto:nakagawa-t at ifrec.osaka-u.ac.jp>> wrote:
>
>
>     Hi,
>
>     is the baselining limited to certain kinds of channels?
>     I can't get it to work right for physiology (emg,ecg), the means
>     aren't
>     0 for the baseline period.
>
>     Here the details:
>     I have this raw file:
>     https://www.dropbox.com/s/6t4z52tno5tl228/raw.fif?dl=0   (1GB!)
>     and I create epochs with the below code, baseline correcting it.
>
>     It may also be related to a possible bug or something I'm doing
>     wrong in
>     pandas dataframe indexing, see here:
>     http://stackoverflow.com/questions/36589702/pandas-dataframe-row-multiindex-skip-one/36590006#36590006
>
>     Thanks for any input.
>     best,
>     Tristan
>
>     -----------------------------------
>
>     import pandas as pd
>     import numpy as np
>     import matplotlib.pyplot as plt
>     import mne
>     import scipy.signal as signal
>
>     # load data
>     raw = Raw('raw.fif', preload = True) #load Data
>     raw._first_samps[0]=0 # From my older question, I still need this..
>     mne.channels.rename_channels(raw.info
>     <http://raw.info>,{'BIO003':'emg1','BIO004':'emg2',
>                                          
>      'BIO005':'emg3','BIO006':'emg4',
>                                          
>      'BIO007':'emg5','BIO008':'ecg'})
>     raw.set_channel_types({'emg1':'emg','emg2':'emg','emg3':'emg','emg4':'emg','emg5':'emg','ecg':'ecg'})
>     picks_emg = mne.pick_types(raw.info <http://raw.info>,emg=True)
>     # preprocess emg signals: hilbert envelope!
>     raw.filter(10, 150, picks=picks_emg)
>
>     # For viewability, I put the events below, load them first!
>
>     # Create & save Epochs
>     event_id = dict(CS=1, OS=2, NS=3)
>     tmin = -1.1  # start of each epoch
>     tmax = 6 # end of each epoch
>     baseline = (-1, -.05) # in seconds
>     epochs = mne.Epochs(raw,events, event_id, tmin=tmin, tmax=tmax,
>     proj=False, picks=[0,1,2,3,4,5],
>                         baseline=(-.5,.05), preload=True,reject=None)
>
>     # Confirm that mean for baseline period is 0:
>     test=epochs['CS'].to_data_frame()
>     # mean=test.loc[idx['CS',:, -1000:-50], :].mean() # doesn't work, see
>     stackoverflow link!
>     mean=test.loc[idx['CS',test.index.levels[1], -1000:-50], :].mean()
>     print(mean)
>
>     # events
>     events=np.array([[ 11527,      0,      1],
>            [ 22344,      0,      2],
>            [ 31541,      0,      3],
>            [ 42542,      0,      2],
>            [ 52757,      0,      3],
>            [ 61972,      0,      1],
>            [ 71586,      0,      3],
>            [ 81785,      0,      3],
>            [ 92786,      0,      3],
>            [103803,      0,      3],
>            [113000,      0,      3],
>            [124001,      0,      2],
>            [135018,      0,      2],
>            [145016,      0,      2],
>            [155616,      0,      2],
>            [164830,      0,      3],
>            [174629,      0,      2],
>            [185629,      0,      1],
>            [196245,      0,      3],
>            [207246,      0,      2],
>            [217645,      0,      1],
>            [226659,      0,      1],
>            [237459,      0,      2],
>            [248476,      0,      1],
>            [258875,      0,      1],
>            [269475,      0,      1],
>            [280091,      0,      1],
>            [289890,      0,      3],
>            [300289,      0,      1],
>            [309503,      0,      2],
>            [319903,      0,      1],
>            [328900,      0,      1],
>            [338515,      0,      3],
>            [347512,      0,      2],
>            [356726,      0,      2],
>            [367526,      0,      3],
>            [377926,      0,      2],
>            [387540,      0,      2],
>            [398557,      0,      3],
>            [407571,      0,      3],
>            [417370,      0,      1],
>            [427168,      0,      2],
>            [437784,      0,      1],
>            [448384,      0,      3],
>            [457782,      0,      1],
>            [467797,      0,      1],
>            [477596,      0,      1],
>            [488012,      0,      2],
>            [498411,      0,      3],
>            [509011,      0,      3],
>            [518626,      0,      2],
>            [529025,      0,      3],
>            [539425,      0,      3],
>            [548639,      0,      2],
>            [557836,      0,      2],
>            [567835,      0,      3],
>            [578852,      0,      1],
>            [588450,      0,      1],
>            [598666,      0,      1],
>            [608064,      0,      2]])
>
>
>
>     _______________________________________________
>     Mne_analysis mailing list
>     Mne_analysis at nmr.mgh.harvard.edu
>     <mailto:Mne_analysis at nmr.mgh.harvard.edu>
>     https://mail.nmr.mgh.harvard.edu/mailman/listinfo/mne_analysis
>
>
>     The information in this e-mail is intended only for the person to
>     whom it is
>     addressed. If you believe this e-mail was sent to you in error and
>     the e-mail
>     contains patient information, please contact the Partners
>     Compliance HelpLine at
>     http://www.partners.org/complianceline . If the e-mail was sent to
>     you in error
>     but does not contain patient information, please contact the
>     sender and properly
>     dispose of the e-mail.
>
>

-- 
Tristan T. Nakagawa, Ph.D.
Laboratory for Brain-Immune Interaction,
iFReC, Osaka University 
3-1 Yamadaoka, Suita, Osaka, Japan
Tel: 0668789710
Office: CiNet R 2B6-2
http://seymourlab.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20160415/493c0b0b/attachment-0001.html 


More information about the Mne_analysis mailing list