U f{j6@sddlmZddlZddlmZddlZddlm Z m Z m Z m Z ddl mZddlmZmZmZmZmZmZmZmZddlmZmZdd lmZmZmZmZm Z m!Z!m"Z"m#Z#dd l$m%Z%dd l&m'Z'm(Z(dd l)m*Z*dd l+m,Z,ddl m-Z-m.Z.m/Z/m0Z0m1Z1m2Z2m3Z3ddlm4Z4ddl5m6Z6ddl7m8Z8ddl9m:Z:ddl;mZ>m?Z?m@Z@mAZAmBZBmCZCmDZDmEZEmFZFmGZGmHZHmIZImJZJmKZKmLZLmMZMmNZNddlOmPZPddlQmRZRmSZSmTZTmUZUmVZVddZWddZXeKGdddeYZZe@d_ddZ[e@d`dd Z\d!d"Z]e@dad(d)Z^dbd+d,Z_e@dcd.d/Z`d0d1Zad2d3Zbd4d5Zcddd6d7Zdd8d9Zeded:d;ZfGdd?d?eRZhe@d&dd@dAdBZidCdDZjdEdFZke@dfdHdIZle@dgdJdKZme@dhdNdOZndidPdQZoe@djdSdTZpe@dkdUdVZqe@dldWdXZrdYdZZse@dmdd\d]d^ZtdS)ndeepcopyN)log)_INTERPOLATION_DEFAULT_EXTRAPOLATE_DEFAULT_BORDER_DEFAULTDEFAULTS)start_and_end_file)make_projector _proj_equal activate_proj _check_projs_needs_eeg_average_ref_proj_has_eeg_average_ref_proj _read_proj _write_proj) fiff_openRawArray) pick_typespick_channels_cov pick_channels pick_info_picks_by_type_pick_data_channels _picks_to_idx_DATA_CH_TYPES_SPLIT)FIFF)_read_bad_channels create_info)find_tag) dir_tree_find) start_block end_block write_intwrite_name_list write_doublewrite_float_matrix write_string)_handle_default)Epochs)make_fixed_length_events) EvokedArray) compute_rank) check_fnameloggerverbose check_version _time_maskwarncopy_function_doc_to_method_doc_pl_undo_scaling_cov _scaled_array_validate_type _check_optioneighfill_doc _on_missing_check_on_missing _check_fname)viz) BaseEstimatorEmpiricalCovariance_logdetempirical_covariancelog_likelihoodcCsL|j|jkrtddd|dD}dd|dD}||krHtddS)Nz6Both Covariance do not have the same list of channels.cSsg|] }t|qSstr.0crErEF/var/www/html/bwave/python/venv/lib/python3.8/site-packages/mne/cov.py 3sz'_check_covs_algebra..projscSsg|] }t|qSrErFrHrErErKrL4sz=Both Covariance do not have the same list of SSP projections.)ch_names ValueError)Zcov1Zcov2Zprojs1Zprojs2rErErK_check_covs_algebra/s  rPcCsht|j|||jdd}|dk r2t|ddnd}|dk rTt|dddnd}t||d}|S)zGet the slice.sfreq)rQNrr)r2timesinfonpwhereslice)epochstmintmaxmaskZtstarttendtslicerErErK _get_tslice:s " r^c@seZdZdZed*ddddZeddZedd Zed d Z ed dd ddZ ddZ ddZ ddZ ddZddZddZddZeeejjgdd dddfdd Zeddddddddd!d"d#d dd dddd$d%edededdfd&d'Zd+d(d)ZdS), CovarianceafNoise covariance matrix. .. warning:: This class should not be instantiated directly, but instead should be created using a covariance reading or computation function. Parameters ---------- data : array-like The data. names : list of str Channel names. bads : list of str Bad channels. projs : list Projection vectors. nfree : int Degrees of freedom. eig : array-like | None Eigenvalues. eigvec : array-like | None Eigenvectors. method : str | None The method used to compute the covariance. loglik : float The log likelihood. %(verbose)s Attributes ---------- data : array of shape (n_channels, n_channels) The covariance. ch_names : list of str List of channels' names. nfree : int Number of degrees of freedom i.e. number of time points used. dim : int The number of channels ``n_channels``. See Also -------- compute_covariance compute_raw_covariance make_ad_hoc_cov read_cov Nr0c  CsZ|jdk} t|}|j|t||||||| |tjd |dk rF||d<| dk rV| |d<dS)zInit of covariance.r) datadimnamesbadsnfreeeigeigvecdiagrMkindNmethodloglik)ndimrupdatelenrFIFFV_MNE_NOISE_COV) selfrarcrdrMrerfrgrjrkr0rhrErErK__init__ts zCovariance.__init__cCs|dS)z'Numpy array of Noise covariance matrix.rarErprErErKraszCovariance.datacCs|dS)zChannel names.rcrErrrErErKrNszCovariance.ch_namescCs|dS)zNumber of degrees of freedom.rerErrrErErKreszCovariance.nfreeF overwriter0c Cs:t|ddt||d}t|}t||W5QRXdS)zSave covariance matrix in a FIF file. Parameters ---------- fname : str Output filename. %(overwrite)s .. versionadded:: 1.0 %(verbose)s covariancez-cov.fifz -cov.fif.gzz_cov.fifz _cov.fif.gz)fnamertN)r.r>r _write_cov)rprwrtr0fidrErErKsaves   zCovariance.savecCst|S)zCopy the Covariance object. Returns ------- cov : instance of Covariance The copied object. rrrrErErKcopyszCovariance.copycCs:|dr |Sd|d<t|d|d<d|d<d|d<|S)aJSet covariance to be processed as being diagonal. Returns ------- cov : dict The covariance. Notes ----- This function allows creation of inverse operators equivalent to using the old "--diagnoise" mne option. This function operates in place. rhTraNrfrg)rUrhrrrErErKas_diagszCovariance.as_diagcCs$|dr d|d<|d|d<|S)NrhF)r|rrrErErK _as_squares zCovariance._as_squarecCsJ|d|jjdkkr,td|jj|df|dr@t|jS|jS)NrhrzOCovariance attributes inconsistent, got data with dimensionality %d but diag=%s)rarl RuntimeErrorrUrhr{rrrErErK _get_squareszCovariance._get_squarecCsJ|jjdkrd|jj}n d|jj}|d|j7}|d|j7}d|S)Nzsize : %s x %sz diagonal : %sz, n_samples : %sz , data : %sz)rarlshapesizere)rpsrErErK__repr__s   zCovariance.__repr__cCs|t|||}|d|d|d|d|d|d|d<|d|d7<tt|d|d|d<|S)@Add Covariance taking into account number of degrees of freedom.rarerd)rPr{listsetunion)rpcovZthis_covrErErK__add__s zCovariance.__add__cCs|t|||d|d|d|d|d|d|ddd<|d|d7<tt|d|d|d<|S)rrareNrd)rPrrr)rprrErErK__iadd__s zCovariance.__iadd__Tc Cstj||||||||SN)r?miscplot_cov)rprTexcludecolorbarprojZshow_svdshowr0rErErKplotszCovariance.plot@rz%3.1fheadc Csddlm}|||dd\}}}}tt|ddtjf|}|dk rptt|||}| dkrdd} |dkrpd}| dkrdd td D} |dkrd d td D}|j d g|||||||| | | | | ||||||||||||ddS)a.Plot a topomap of the covariance diagonal. Parameters ---------- %(info_not_none)s %(ch_type_topomap)s %(vmin_vmax_topomap)s %(cmap_topomap)s %(sensors_topomap)s %(colorbar_topomap)s %(scalings_topomap)s %(units_topomap)s %(res_topomap)s %(size_topomap)s %(cbar_fmt_topomap)s %(proj_plot)s %(show)s %(show_names_topomap)s %(title_none)s %(mask_topomap)s %(mask_params_topomap)s %(outlines_topomap)s %(contours_topomap)s %(image_interp_topomap)s %(axes_topomap)s %(extrapolate_topomap)s %(sphere_topomap_auto)s %(border_topomap)s noise_cov : instance of Covariance | None If not None, whiten the instance with ``noise_cov`` before plotting. %(verbose)s Returns ------- fig : instance of Figure The matplotlib figure. Notes ----- .. versionadded:: 0.21 r)_index_info_covrErNZAU?cSsi|]\}}|d|dqS)(u)²rErIkvrErErK >sz+Covariance.plot_topomap..unitscSsi|]\}}|||qSrErErrErErKr@sscalingsr)rSch_typevminvmaxcmapsensorsrrrresrcbar_fmtrr show_namestitler[ mask_paramsoutlinescontours image_interpaxes extrapolatesphereborder time_format) Zviz.miscrr,rUrhnewaxis whiten_evokedr items plot_topomap) rprTrrrrrrrrrrrrrrrr[rrrrrrrr noise_covr0rC_evokedrErErKrsR4 zCovariance.plot_topomapcCst||g|ddS)aPick channels from this covariance matrix. Parameters ---------- ch_names : list of str List of channels to keep. All other channels are dropped. ordered : bool If True (default False), ensure that the order of the channels matches the order of ``ch_names``. Returns ------- cov : instance of Covariance. The modified covariance matrix. Notes ----- Operates in-place. .. versionadded:: 0.20.0 F)rorderedr{)r)rprNrrErErKrKs zCovariance.pick_channels)NNNN)F)__name__ __module__ __qualname____doc__r0rqpropertyrarNrerzr{r|r}rrrrr4r?rrrrrrrrrErErErKr_Csr/          Lr_c Csbt|ddt|ddd}t|dd\}}|&}tft||tjddW5QRSQRXdS) aRead a noise covariance from a FIF file. Parameters ---------- fname : str The name of file containing the covariance matrix. It should end with -cov.fif or -cov.fif.gz. %(verbose)s Returns ------- cov : Covariance The noise covariance matrix. See Also -------- write_cov, compute_covariance, compute_raw_covariance rurvTread)rwZ must_existrtNr)limited)r.r>rr_ _read_covrro)rwr0ftreeryrErErKread_covhs rc stdddd}td|}tt|}tdd|d|d|d fD],\}}}t||d }|||t||<qFfd d |D} t|| d dddS)uCreate an ad hoc noise covariance. Parameters ---------- %(info_not_none)s std : dict of float | None Standard_deviation of the diagonal elements. If dict, keys should be ``'grad'`` for gradiometers, ``'mag'`` for magnetometers and ``'eeg'`` for EEG channels. If None, default values will be used (see Notes). %(verbose)s Returns ------- cov : instance of Covariance The ad hoc diagonal noise covariance for the M/EEG data channels. Notes ----- The default noise values are 5 fT/cm, 20 fT, and 0.2 µV for gradiometers, magnetometers, and EEG channels respectively. .. versionadded:: 0.9.0 TrE)megeegrZ noise_std)gradmagF)FFTrrr)rrcsg|]}d|qSrNrE)rIpickrTrErKrLsz#make_ad_hoc_cov..rdrMrre)rr)rUzerosrnzip searchsortedr_) rTZstdr0picksrarrvalZ these_picksrNrErrKmake_ad_hoc_covs rcCs<d|dd}|dkr td||kr8td||fdS)z?Check to see if there are enough samples for reliable cov calc. rrrz1No samples found to compute the covariance matrixzOToo few samples (required : %d got : %d), covariance estimate may be unreliableN)rOr3) n_samplesn_chanZ n_samples_minrErErK_check_n_samplessr皙? empiricalFTcsl|dkr dnt|}djd}|dkr8jd|nt|}|dkrP||nt|}||}td|||}tdt|t|f|dkrtjd}t |t jd d }nt d}t j|}t |dd |d|||d d d | d }|dkrd }t|tr |d kr ||}d }d }d }|D]>}||}||jdd7}|t||j7}||jd7}q*t|t|||dddf|dddf|8}||d}td|tdfdd|DfddjdD}t||jd|ddS~~||jjd djdd}|j|tjddtjf8_d|_t|d||| | | | |d S)aEstimate noise covariance matrix from a continuous segment of raw data. It is typically useful to estimate a noise covariance from empty room data or time intervals before starting the stimulation. .. note:: To estimate the noise covariance from epoched data, use :func:`mne.compute_covariance` instead. Parameters ---------- raw : instance of Raw Raw data. tmin : float Beginning of time interval in seconds. Defaults to 0. tmax : float | None (default None) End of time interval in seconds. If None (default), use the end of the recording. tstep : float (default 0.2) Length of data chunks for artifact rejection in seconds. Can also be None to use a single epoch of (tmax - tmin) duration. This can use a lot of memory for large ``Raw`` instances. reject : dict | None (default None) Rejection parameters based on peak-to-peak amplitude. Valid keys are 'grad' | 'mag' | 'eeg' | 'eog' | 'ecg'. If reject is None then no rejection is done. Example:: reject = dict(grad=4000e-13, # T / m (gradiometers) mag=4e-12, # T (magnetometers) eeg=40e-6, # V (EEG channels) eog=250e-6 # V (EOG channels) ) flat : dict | None (default None) Rejection parameters based on flatness of signal. Valid keys are 'grad' | 'mag' | 'eeg' | 'eog' | 'ecg', and values are floats that set the minimum acceptable peak-to-peak amplitude. If flat is None then no rejection is done. %(picks_good_data_noref)s method : str | list | None (default 'empirical') The method used for covariance estimation. See :func:`mne.compute_covariance`. .. versionadded:: 0.12 method_params : dict | None (default None) Additional parameters to the estimation procedure. See :func:`mne.compute_covariance`. .. versionadded:: 0.12 cv : int | sklearn.model_selection object (default 3) The cross validation method. Defaults to 3, which will internally trigger by default :class:`sklearn.model_selection.KFold` with 3 splits. .. versionadded:: 0.12 scalings : dict | None (default None) Defaults to ``dict(mag=1e15, grad=1e13, eeg=1e6)``. These defaults will scale magnetometers and gradiometers at the same unit. .. versionadded:: 0.12 %(n_jobs)s .. versionadded:: 0.12 return_estimators : bool (default False) Whether to return all estimators or the best. Only considered if method equals 'auto' or is a list of str. Defaults to False. .. versionadded:: 0.12 %(reject_by_annotation_epochs)s .. versionadded:: 0.14 %(rank_none)s .. versionadded:: 0.17 .. versionadded:: 0.18 Support for 'info' mode. %(verbose)s Returns ------- cov : instance of Covariance | list The computed covariance. If method equals 'auto' or is a list of str and return_estimators equals True, a list of covariance estimators is returned (sorted by log-likelihood, from high to low, i.e. from best to worst). See Also -------- compute_covariance : Estimate noise covariance matrix from epoched data. Notes ----- This function will: 1. Partition the data into evenly spaced, equal-length epochs. 2. Load them into memory. 3. Subtract the mean across all time points and epochs for each channel. 4. Process the :class:`Epochs` by :func:`compute_covariance`. This will produce a slightly different result compared to using :func:`make_fixed_length_events`, :class:`Epochs`, and :func:`compute_covariance` directly, since that would (with the recommended baseline correction) subtract the mean across time *for each epoch* (instead of across epochs) for each channel. NrrQrRrzUsing up to %s segment%sZnchanFZ with_ref_megr)baselinerrejectflatr0Zpreloadrreject_by_annotationrZaxisNumber of samples used : %d[done]csg|]}jd|qSrrrIr)rawrErKrLPsz*compute_raw_covariance..csg|]}|kr|qSrErE)rIbrrErKrLQsrdrMr)NNT)keep_sample_meanrj method_paramscvrn_jobsreturn_estimatorsrank)floatrTrSr+r/rnr5rUarangeZin1drrWrr* isinstancerGsumdotTrrr_Z load_data_datameanrrcompute_covariance)rrYrZZtsteprrrrjrrrrrrrr0dtZtstep_m1eventsZ pick_maskrXrarmuZ raw_segmentrdZch_meansrE)rNrrKcompute_raw_covariancesvq  (   rrjc sdddddddddddddddddtddd dddd d id d id }tD]}d|d |<qXt|tr|D]8\}} ||krtd|d|f||||q|| di dd} d| krdksntd| fd} |d krdg}n |dkr$|r$d} dd ddg}t|t t fs:|g}t fdd|Dsftdj ||dt|tr||dks| r||d|D] } | dkrtd| |fq|st|dksd|krtd|f|D]$\} }| dd dkrtd q||fS)!z/Check that method and method_params are usable.) autordiagonal_fixed ledoit_wolfoasshrunkpcafactor_analysis shrinkageFTstore_precisionassume_centered皙?)rrrriter_n_componentsN)rrrrrrrrrzkey (%s) must be "%s"z" or "rrz)shrinkage must be between 0 and 1, got %srrrrc3s|]}|kVqdSrrEraccepted_methodsrErK sz'_check_method_params..zoInvalid {name} ({method}). Accepted values (individually or in a list) are any of "{accepted_methods}" or None.)namerjr full)rrz8%s can so far only be used with rank="full", got rank=%rz=`keep_sample_mean=False` is only supportedwith %s="empirical"rz=`assume_centered` must be True if `keep_sample_mean` is False)rUZlogspacerrdictrrOjoinrmgetrtupleallformatrGpopindexrn)rjrrr allow_autor_method_paramsrkeyvaluesrZwas_automethod_prrEr rK_check_method_paramsds|         rraisec+ st|}t||| d\}}~ddtts:ntfddDgtfddDrptdd jd }t| d t D]j\}}|j |d k|jd d kks|d k rt |d |jd d sd|||jd f}t | |d qd jd}|d krd jd}dd D]P}|jd jkrLtdt|jd|D]\}}t||s\tdq\q.t|}d j}dd D]2}|jd|krtd|j|krtdqtd j}t dd|Dt fddD}d j}st}d g|}t j|t jd}t j|t jd}t D]j\}}t||||D]N}|f}s|||7<|||jd7<||d7<qrqZ||} t | |d}!ddt||D}t|}td ||fddDt tt|}tdkrTt d nd t jd}"t|"tj t!||||| d||| d }#dkr|#dd }$|$|"d9}$|D]}%|$|%8}$q|$|!}$t}&|#"D]H\}'}(t#|($d ||d||"dd!}$|$j%fd"|'i|(|&&|$qt'd#|"|&jd$d%dd&t|&dkrd'g}|&D] })|&d(|)d"|)d)fqnt'd*(|| r|&}*n|&d }*t'd+)|*d"n|&d }*t'd,|*S)-aEstimate noise covariance matrix from epochs. The noise covariance is typically estimated on pre-stimulus periods when the stimulus onset is defined from events. If the covariance is computed for multiple event types (events with different IDs), the following two options can be used and combined: 1. either an Epochs object for each event type is created and a list of Epochs is passed to this function. 2. an Epochs object is created for multiple events and passed to this function. .. note:: To estimate the noise covariance from non-epoched raw data, such as an empty-room recording, use :func:`mne.compute_raw_covariance` instead. Parameters ---------- epochs : instance of Epochs, or list of Epochs The epochs. keep_sample_mean : bool (default True) If False, the average response over epochs is computed for each event type and subtracted during the covariance computation. This is useful if the evoked response from a previous stimulus extends into the baseline period of the next. Note. This option is only implemented for method='empirical'. tmin : float | None (default None) Start time for baseline. If None start at first sample. tmax : float | None (default None) End time for baseline. If None end at last sample. projs : list of Projection | None (default None) List of projectors to use in covariance calculation, or None to indicate that the projectors from the epochs should be inherited. If None, then projectors from all epochs must match. method : str | list | None (default 'empirical') The method used for covariance estimation. If 'empirical' (default), the sample covariance will be computed. A list can be passed to perform estimates using multiple methods. If 'auto' or a list of methods, the best estimator will be determined based on log-likelihood and cross-validation on unseen data as described in :footcite:`EngemannGramfort2015`. Valid methods are 'empirical', 'diagonal_fixed', 'shrunk', 'oas', 'ledoit_wolf', 'factor_analysis', 'shrinkage', and 'pca' (see Notes). If ``'auto'``, it expands to:: ['shrunk', 'diagonal_fixed', 'empirical', 'factor_analysis'] ``'factor_analysis'`` is removed when ``rank`` is not 'full'. The ``'auto'`` mode is not recommended if there are many segments of data, since computation can take a long time. .. versionadded:: 0.9.0 method_params : dict | None (default None) Additional parameters to the estimation procedure. Only considered if method is not None. Keys must correspond to the value(s) of ``method``. If None (default), expands to the following (with the addition of ``{'store_precision': False, 'assume_centered': True} for all methods except ``'factor_analysis'`` and ``'pca'``):: {'diagonal_fixed': {'grad': 0.1, 'mag': 0.1, 'eeg': 0.1, ...}, 'shrinkage': {'shrikage': 0.1}, 'shrunk': {'shrinkage': np.logspace(-4, 0, 30)}, 'pca': {'iter_n_components': None}, 'factor_analysis': {'iter_n_components': None}} cv : int | sklearn.model_selection object (default 3) The cross validation method. Defaults to 3, which will internally trigger by default :class:`sklearn.model_selection.KFold` with 3 splits. scalings : dict | None (default None) Defaults to ``dict(mag=1e15, grad=1e13, eeg=1e6)``. These defaults will scale data to roughly the same order of magnitude. %(n_jobs)s return_estimators : bool (default False) Whether to return all estimators or the best. Only considered if method equals 'auto' or is a list of str. Defaults to False. on_mismatch : str What to do when the MEG<->Head transformations do not match between epochs. If "raise" (default) an error is raised, if "warn" then a warning is emitted, if "ignore" then nothing is printed. Having mismatched transforms can in some cases lead to unexpected or unstable results in covariance calculation, e.g. when data have been processed with Maxwell filtering but not transformed to the same head position. %(rank_none)s .. versionadded:: 0.17 .. versionadded:: 0.18 Support for 'info' mode. %(verbose)s Returns ------- cov : instance of Covariance | list The computed covariance. If method equals 'auto' or is a list of str and return_estimators equals True, a list of covariance estimators is returned (sorted by log-likelihood, from high to low, i.e. from best to worst). See Also -------- compute_raw_covariance : Estimate noise covariance from raw data, such as empty-room recordings. Notes ----- Baseline correction or sufficient high-passing should be used when creating the :class:`Epochs` to ensure that the data are zero mean, otherwise the computed covariance matrix will be inaccurate. Valid ``method`` strings are: * ``'empirical'`` The empirical or sample covariance (default) * ``'diagonal_fixed'`` A diagonal regularization based on channel types as in :func:`mne.cov.regularize`. * ``'shrinkage'`` Fixed shrinkage. .. versionadded:: 0.16 * ``'ledoit_wolf'`` The Ledoit-Wolf estimator, which uses an empirical formula for the optimal shrinkage value :footcite:`LedoitWolf2004`. * ``'oas'`` The OAS estimator :footcite:`ChenEtAl2010`, which uses a different empricial formula for the optimal shrinkage value. .. versionadded:: 0.16 * ``'shrunk'`` Like 'ledoit_wolf', but with cross-validation for optimal alpha. * ``'pca'`` Probabilistic PCA with low rank :footcite:`TippingBishop1999`. * ``'factor_analysis'`` Factor analysis with low rank :footcite:`Barber2012`. ``'ledoit_wolf'`` and ``'pca'`` are similar to ``'shrunk'`` and ``'factor_analysis'``, respectively, except that they use cross validation (which is useful when samples are correlated, which is often the case for M/EEG data). The former two are not included in the ``'auto'`` mode to avoid redundancy. For multiple event types, it is also possible to create a single :class:`Epochs` object with events obtained using :func:`mne.merge_events`. However, the resulting covariance matrix will only be correct if ``keep_sample_mean is True``. The covariance can be unstable if the number of samples is small. In that case it is common to regularize the covariance estimate. The ``method`` parameter allows to regularize the covariance in an automated way. It also allows to select between different alternative estimation algorithms which themselves achieve regularization. Details are described in :footcite:`EngemannGramfort2015`. For more information on the advanced estimation methods, see :ref:`the sklearn manual `. References ---------- .. footbibliography:: )rcs.tjdkr$fddjDngS)Nrcsg|] }|qSrErErrXrErKrL]sz>compute_covariance.._unpack_epochs..rnZevent_idr rEr rK_unpack_epochs[sz*compute_covariance.._unpack_epochscsg|] }|qSrErE)rIepoch)r"rErKrLesz&compute_covariance..c3s*|]"}|jdko |jddko VqdS)NZhighpass?)rrT)rIepochs_t)rrErKr hsz%compute_covariance..zFEpochs are not baseline corrected, covariance matrix may be inaccuraterZ dev_head_t on_mismatchNZtranszGMEG<->Head transform mismatch between epochs[0]: %s and epochs[%s]: %srdrMrz+Epochs must agree on the use of projectionsz Epochs must have same projectorsz"Epochs must have same bad channelsz#Epochs must have same channel namescSsg|] \}}|qSrErE)rIrrrErErKrLscsg|]}dj|qSrrrr rErKrLsdtypecSs&g|]\}}d|t||jqS)r)rUrr)rIZn_epochrrErErKrLscs g|]}|jddfqS))r.)get_data)rIee) picks_meegr]rErKrLsrRT rjrrTrr stop_early picks_listrrFrrarrjrcSs|dS)NrkrE)rJrErErKz$compute_covariance..)rreversez1log-likelihood on unseen data (descending order):z %s: %0.3frkz zselecting best estimator: {}r)*_check_scalings_userrrrranyr3rTr= enumerate_check_consistencyrUZallcloser<rrOrr rrNrZ concatenatesortrnrZint64r^rrrZhstackrr_compute_covariance_autorr_rrmappendr/rr)+rXrrYrZrMrjrrrrrr&rr0rorigeir#msgrdr%Zproj_aZproj_brNr/rTZ n_epoch_typesZ data_meanrZn_epochsiieZn_samples_epochZ norm_constZ n_samples_totZcov_datarZmean_covcovsZ this_methodrarJoutrE)r"rXrr,r]rKrs.                          rcCsbt|tr0|D]\}}td|dddgqn$|dk rTt|tjsTtdt|td|}|S)Nzthe keys in `scalings`rrrz1scalings must be a dict, ndarray, or None, got %sr) rrrr9rUZndarray TypeErrortyper))rrrrErErKr3s  r3cCsJtt|t||j||}t|\}}|j}||fS)z3Compute the subspace from a subset of eigenvectors.)rUeyernrconjrr:)rfrgr[PrErErK_eigvec_subspaces,  rFc / s| } tt|j|ddd| ||} t|j||Jt|j|} t| || ddd\} } }| |} t|| j}t|dfdd|D}tt|krt |n|}t d t|| j dft }d }td }|st|d ks|dd krtd|f|D]}|}t|r*|jn|}t ||||}i}|d krtf|}|||||j|f~n|dkrtfd|i|}|||||j|f~n|dkrRddlm}g}|f|}|D]2\}}||dd|f|||j|fqtfd|i|}|||||j|f~~nB|dkrddlm} g}| f|}!|D]2\}}|!|dd|f|||!j|fqztfd|i|}|||||j|f~!~n|dkrtf|}|||||j|f~nz|dkrddlm}"ddlm }#|!d}$d|$ig}%g}|"|#f||%|d}&|D]4\}}|&|dd|f|||&j"j#|fqlddt$|D}tfd|i|}|||||j|f~$~n|dkr:| dkst%t&||||||d\}'}|'|||'|''|f~'nZ|dkr| dksRt%t&||||||d\}(}|(|||(|('|f~(ntdt d q t|d krt d!t(})t)|D]\}*\}+},}-t|d krt*||+||}.nd}.t| jt|,| },|,|j d|j dd 9},t+|,||||*}t|rN|jn|}t(|.|,|+d"|)|<|)|,|-qW5QRX|)S)#zCompute covariance auto mode.NF)r{r0T) proj_subspacedo_compute_rankrcs g|]\}}|t|fqSrE)rUr)rIrrusedrErKrLsz,_compute_covariance_auto..z Reducing data rank from %s -> %szEstimating covariance using %sZsklearnrrzCscikit-learn is not installed, `method` must be `empirical`, got %srrTr) LedoitWolfrr)OASr) GridSearchCV)ShrunkCovariance)rcSsg|] }|dqSr'rErHrErErKrLQsrr)rrrr.rz1Oh no! Your estimator does not have a .fit methodzDone.z4Using cross-validation to select the best estimator.)rkra estimator)-r-rrr7rUr _smart_eighrVrnrr/rTrrr1rOr{callablerupperrAfitr9 covariance__RegCovariancesklearn.covariancerKZ shrinkage__ShrunkCovariancerLsklearn.model_selectionrMrNrZbest_estimator_rrAssertionError_auto_low_rank_modelZget_covariancerr5 _cross_valr6rm)/rarjrTrrrrr.r/rZ orig_rankrrrgr[Zsub_picks_listZsub_infoZestimator_cov_infor<Z ok_sklearnrZdata_rmp_infoestrKZ shrinkagesZlwrrscrLrrMrNrZtuned_parametersgsrfar@r;rOr runtime_inforkrErIrKr8s                                         r8cCs^|}|j\}}d|t||jdd}|d|tdtjt|8}t|}|S)z@Compute the Gaussian log likelihood of X under the model in est.grrr$@) get_precisionrrUrrrpirBr)r^Xy precisionrZ n_featuresZlog_liker@rErErK_gaussian_loglik_scorers  " ric Cs$ddlm}t|||||tdS)zCompute cross validation.r)cross_val_score)rrZscoring)rXrjrUrri)rar^rrrjrErErKr[s r[c Cst|}|d}|dkr.td|jdd}ddlm}m} |dkrL| } n|dksXt|} | f|} d| _ tj |tj d } | tj ttt|} | |jdkrtd | |jdft|D]\} }|| _ zt|| ||d }Wntk rtj}YnXt|s|dkr*td qtd ||f|tj krR|| | <| dkrtt| | d| dkr|rtdqqt| rtdt| }||}| _ td|t|| ||d}| |fS)zCompute latent variable models.r Nrr)PCAFactorAnalysisrrr(zDYou are trying to estimate %i components on matrix with %i features.)rar^rrz4... infinite values encountered. stopping estimationz... rank: %i - loglik: %0.3frz early stopping parameter search.zkOh no! Could not estimate covariance because all scores were NaN. Please contact the MNE-Python developers.z... best model at rank = %i)Zranksscoresbestr)rrrUrrZsklearn.decompositionrlrmrYZ n_componentsZ empty_likefloat64fillnanmaxrr3r5r[rOinfisinfr/rTrdiffisnanr~Z nanargmaxarray)ramoderrrr.r0r rlrmr^rnZmax_nr=nscoreZi_scorerorbrErErKrZsX       ,  rZc@s4eZdZdZd ddZddZdd d Zd d ZdS)rU Aux class.rFcCsd||_||_||_||_||_||_||_||_||_| |_ | |_ | |_ | |_ | |_ ||_||_dSr)rTrrrseegdbsecoghbohbrfnirs_cw_amplitudefnirs_fd_ac_amplitudefnirs_fd_phasefnirs_odcsdrr)rprTrrrr}rrrrrrrrr~rrrErErKrqs z_RegCovariance.__init__cCst|j|jd|_|j|j|_d|j|jj|_t|j|jd|jd|jdt |jd}t ||jdd|j |j |j |j|j|j|j|jdd }|j|j_|_|S) zt||jjdd}t|jr*d||j<t||j}|S)2Delegate to modified EmpiricalCovariance instance.T)rr)rCrZ location_rUr4rrDrd)rprrgZtest_covrrErErKr{Ds   z_ShrunkCovariance.scorecCs |jS)rrrrrErErKrdNsz_ShrunkCovariance.get_precision)r)NrrErErErKrW s  - rWrscCs|j|||ddS)aAWrite a noise covariance matrix. Parameters ---------- fname : str The name of the file. It should end with -cov.fif or -cov.fif.gz. cov : Covariance The noise covariance matrix. %(overwrite)s .. versionadded:: 1.0 %(verbose)s See Also -------- read_cov rsN)rz)rwrrtr0rErErK write_covVsrcs.tjdkr$fddjDngS)z Aux Function.rcsg|] }|qSrErErr rErKrLrsz"_unpack_epochs..r!r rEr rKr"osr"cCst|dd\}}|j}tt|t}d|d| <d|d| <td||rVdndf|rv|d|  }|||fS) z*Get whitener params for a set of channels.T)Z overwrite_arNFz- Setting small %s eigenvalues to zero (%s)z using PCAz without PCA) r:rDrrUonesrnrr/rTr{)Arrrrfrgr[rErErK_get_ch_whitenerys rignorec Cst}t}|dkr|dn|}|D],} | |jkrF||j| q$|| q$t|rftd||t||} t |t |d|}|d|d} t | |t|dt |d|d| dd| ddd }t||||| ||d \} } }|j| | d |S) aKPrepare noise covariance matrix. Parameters ---------- noise_cov : instance of Covariance The noise covariance to process. %(info_not_none)s (Used to get channel types and bad channels). ch_names : list | None The channel names to be considered. Can be None to use ``info['ch_names']``. %(rank_none)s .. versionadded:: 0.18 Support for 'info' mode. scalings : dict | None Data will be rescaled before rank estimation to improve accuracy. If dict, it will override the following dict (default if None):: dict(mag=1e12, grad=1e11, eeg=1e5) %(on_rank_mismatch)s %(verbose)s Returns ------- cov : instance of Covariance A copy of the covariance with the good channels subselected and parameters updated. NrNz0Not all channels present in noise covariance: %srMrdrerjrk)rarcrdrMrerjrkon_rank_mismatch)rfrg)rrNr9rrnr~rrUrrrr_rrrPrm)rrTrNrrrr0Z noise_cov_idxmissingrJrrMrfrgrrErErKprepare_noise_covs:         rc shtd|}|dkrdn|}|dkr.dn|}d|krVtfdd|Dd|ksftt|} t||\} } } t|tr|d}| dkrtd | t | t || j }t||g|d}|rt ||||d }|j d kr|jd|jd kst|rDt||jdkrDt | t | t | tfS|jt jkrVtnt}t | |}t | | f|}t | t}td dgdD]\}}t|dkrq|t ||}||kr|dkr|d}n||}t|d||\}}}|rt|||\}}|||||<|t ||<||<|dkrtrt|st dq|||fS)z@Compute eigh of C taking into account rank and ch_type scalings.scalings_cov_rankNrMrNcsg|]}d|qSr)rrHrrErKrLsz_smart_eigh..rarz5 Created an SSP operator (subspace dimension = %d)rrrTF) meg_combinedref_megrrrrrzNo average EEG reference present in info["projs"], covariance may be adversely affected. Consider recomputing covariance using with an average eeg reference projector added.)!r)rrYrnr rr_r/rTrUrrr-rlrrrrrCrr)Zcomplex_complexrrrrrrRrFrrr3)rrTrrrMrNrGrHrr0rrncomprrr)rfrgr[rrthis_CZ this_rankr>ZevmrErrKrPsd   &   $ rPrrdc/ sddlm}|}|td|}t||||| | | | | ||d }|dkrTtd|dkrl|d|d}dd tD}|d krd nd }|tt |||d d t | dgdkr|d kr||krtd||ft d||d<n|j||d|d krt||||}|dt}|D] \}}fdd|D||<q(t||d}|jdd |D}tD]D\}}|D](}|||kr|||qtqtd|qt|d}t |ttt |kst|r|d|d} t| } |D]}|}!||}"t |"dkr,q||}#|#dkrPt d|!qt d|!|#f|t|"|"}$t|$jd}%fdd|"D}&|d kr|r&t| |&\}'}(})|(dkr&| |'dddd|( f}%t d|!|(fn4t!|d|&}*t"||*}+t#|$|+|\})},}-|,|-j$}%t%|%j$t%|$|%}$t&t'|$}.|$j(ddt |$d|#|.7<t%|%t%|$|%j$}$|$|t|"|"<qt!|j|d }"||dt|"|"<|S)!a Regularize noise covariance matrix. This method works by adding a constant to the diagonal for each channel type separately. Special care is taken to keep the rank of the data constant. .. note:: This function is kept for reasons of backward-compatibility. Please consider explicitly using the ``method`` parameter in :func:`mne.compute_covariance` to directly combine estimation with regularization in a data-driven fashion. See the `faq `_ for more information. Parameters ---------- cov : Covariance The noise covariance matrix. %(info_not_none)s (Used to get channel types and bad channels). mag : float (default 0.1) Regularization factor for MEG magnetometers. grad : float (default 0.1) Regularization factor for MEG gradiometers. Must be the same as ``mag`` if data have been processed with SSS. eeg : float (default 0.1) Regularization factor for EEG. exclude : list | 'bads' (default 'bads') List of channels to mark as bad. If 'bads', bads channels are extracted from both info['bads'] and cov['bads']. proj : bool (default True) Apply projections to keep rank of data. seeg : float (default 0.1) Regularization factor for sEEG signals. ecog : float (default 0.1) Regularization factor for ECoG signals. hbo : float (default 0.1) Regularization factor for HBO signals. hbr : float (default 0.1) Regularization factor for HBR signals. fnirs_cw_amplitude : float (default 0.1) Regularization factor for fNIRS CW raw signals. fnirs_fd_ac_amplitude : float (default 0.1) Regularization factor for fNIRS FD AC raw signals. fnirs_fd_phase : float (default 0.1) Regularization factor for fNIRS raw phase signals. fnirs_od : float (default 0.1) Regularization factor for fNIRS optical density signals. csd : float (default 0.1) Regularization factor for EEG-CSD signals. dbs : float (default 0.1) Regularization factor for DBS signals. %(rank_none)s .. versionadded:: 0.17 .. versionadded:: 0.18 Support for 'info' mode. scalings : dict | None Data will be rescaled before rank estimation to improve accuracy. See :func:`mne.compute_covariance`. .. versionadded:: 0.17 %(verbose)s Returns ------- reg_cov : Covariance The regularized covariance matrix. See Also -------- mne.compute_covariance r)linalgr) rr}r~rrrrrrrrNz+exclude must be a list of strings or "bads"rdcSsi|] }|gqSrErErIrrErErKrbszregularize..rrF)rrrrzlOn data where magnetometers and gradiometers are dependent (e.g., SSSed data), mag (%s) must equal grad (%s)z!Regularizing MEG channels jointlyrrNcsg|] }|qSrErE)rIi) info_ch_namesrErKrLuszregularize..)includercSsi|] }|gqSrErErrErErKr}szchannel %s is unknown typerarMrz %s regularization : Nonez %s regularization : %scsg|] }|qSrErErrrErKrLsz3 Created an SSP operator for %s (dimension = %d)rr))scipyrr{r6r)rrOrrmrrnrr/rTr-rrrNr5r9 ExceptionrmaprrYr rRrUrrCrr ZsvdrrrPrrrrhr)/rrTrrrrrr}rrrrrrrrr~rrr0rregsZ picks_dictrZch_names_by_typerZ picks_typeZcov_goodZidx_covrchrrMdescidxregrUZ this_ch_namesrErrZ this_picksZ this_inforgr[sigmarE)rNrrKrsN               $rc Cst|tddf|dkrd}nDt|tsbt|}|dk rJtdt|ftt|dddd}d }t||d d|d \}}|dkrt|j d d dn|}t |}t dd}t |j |||ddd|||d |d}|S)aCompute a regularized covariance from data using sklearn. This is a convenience wrapper for mne.decoding functions, which adopted a slightly different covariance API. Returns ------- cov : ndarray, shape (n_channels, n_channels) The covariance matrix. numericNrz6If reg is a float, method_params must be None (got %s)TF)rrrrrr)rrrg@@rrr-ra)r8rGrrrOrBrrrrrr)r8r) rarrrTrrjr/rrrErErK_regularized_covariancesT     rr3c st|ttfdd} || kr,td| |fdkrRd|krDtdt|d} n6t|dd }fd d |D} ~t|| |||d }t| } | t|dkst|d } | d k}d| |<|dj j dkrt j }nt j}t | df|}dt | |||d f<||d}t | |dj}|}td||d|f|dkrt||}|dd|f}n0|dkrt |dj|}t ||d}|| f}|r||f7}|r||f7}|S)aCompute whitening matrix. Parameters ---------- noise_cov : Covariance The noise covariance. %(info)s Can be None if ``noise_cov`` has already been prepared with :func:`prepare_noise_cov`. %(picks_good_data_noref)s %(rank_none)s .. versionadded:: 0.18 Support for 'info' mode. scalings : dict | None The rescaling method to be applied. See documentation of ``prepare_noise_cov`` for details. return_rank : bool If True, return the rank used to compute the whitener. .. versionadded:: 0.15 pca : bool | str Space to project the data into. Options: :data:`python:True` Whitener will be shape (n_nonzero, n_channels). ``'white'`` Whitener will be shape (n_channels, n_channels), potentially rank deficient, and have the first ``n_channels - n_nonzero`` rows and columns set to zero. :data:`python:False` (default) Whitener will be shape (n_channels, n_channels), potentially rank deficient, and rotated back to the space of the original data. .. versionadded:: 0.18 return_colorer : bool If True, return the colorer as well. %(on_rank_mismatch)s %(verbose)s Returns ------- W : ndarray, shape (n_channels, n_channels) or (n_nonzero, n_channels) The whitening matrix. ch_names : list The channel names. rank : int Rank reduction of the whitener. Returned only if return_rank is True. colorer : ndarray, shape (n_channels, n_channels) or (n_channels, n_nonzero) The coloring matrix. space)TwhiteFzspace must be one of %s, got %sNrfzWinfo can only be None if the noise cov has already been prepared with prepare_noise_covrcFrcsg|]}d|qSrrErrrErKrL!sz$compute_whitener..rrrrgrJrrzd Created the whitener using a noise covariance matrix with rank %d (%d small eigenvalues omitted)rbT)r8rGrrOrrrrnrYr{r)rirUZ complex128rprrrDrrr/rTr)rrTrrrZ return_rankrZreturn_colorerrr0Z _valid_pcasrNrrfZnzeror)WrZn_nzeror@rErrKcompute_whitenersb7       rc Cs`|}t|j|}|r |}t||j|||d\}}t|jt||j ||j |<|S)aSWhiten evoked data using given noise covariance. Parameters ---------- evoked : instance of Evoked The evoked data. noise_cov : instance of Covariance The noise covariance. %(picks_good_data)s diag : bool (default False) If True, whiten using only the diagonal of the covariance. %(rank_none)s .. versionadded:: 0.18 Support for 'info' mode. scalings : dict | None (default None) To achieve reliable rank estimation on multiple sensors, sensors have to be rescaled. This parameter controls the rescaling. If dict, it will override the following default dict (default if None): dict(mag=1e12, grad=1e11, eeg=1e5) %(verbose)s Returns ------- evoked_white : instance of Evoked The whitened evoked data. )rrr) r{rrTr|rrUrZnaverra) rrrrhrrr0rrrErErKrNs    $rc Cs6ddlm}t|tj}t|dkr,tdtt|D]}t|||tj }|dk r8t |j |kr8||} t|| tj }|dkrtdt |j } t|| tj }|dkrd} n t |j } t|| tj}|dkrd} n|j } t|| tj}|dkrd} n |j d} t|| tj}|dkr"g}n"|j d}t|| krDtdt|| tj}|dkrt|| tj}|dkr~td n|j }d }td | | |fn||j s|j }t| | f}||tt| | fdk<||j}|jdd| d d <d}td| | |fnd}|j }td| | |ft|| tj}t|| tj}|dk rx|dk rx|j }|j }nd}d}t|| }t || d}| t|kst!|j"|rd ndkst!t#||| ||||| ||d }| dk r| |d<| dk r| |d<|r|d=|d=|d=|Sq8td|dS)zRead a noise covariance matrix.r)sparsezNo covariance matrices foundNz%Covariance matrix dimension not foundrR:z:Number of names does not match covariance matrix dimensionzNo covariance matrix data foundTz2 %d x %d diagonal covariance (kind = %d) found.rrcFz. %d x %d full covariance (kind = %d) found.z0 %d x %d sparse covariance (kind = %d) found.r) rirhrbrcrarMrdrerfrgrkrjrirbrhz: Did not find the desired covariance matrix (kind = %d))$rrr!r FIFFB_MNE_COVrnrOranger FIFF_MNE_COV_KINDintraFIFF_MNE_COV_DIMFIFF_MNE_COV_NFREEFIFF_MNE_COV_METHODFIFF_MNE_COV_SCOREFIFF_MNE_ROW_NAMESsplit FIFF_MNE_COVFIFF_MNE_COV_DIAGr/rTissparserUrtrilrrrFIFF_MNE_COV_EIGENVALUESFIFF_MNE_COV_EIGENVECTORSrrrYrlr)rynodeZcov_kindrr0rr?rtagthisrbrerjr{rcrarhvalsZtag1Ztag2rfrgrMrdrrErErKr{s                  rcCst|tjt|tj|dt|tj|d|ddkrNt|tj|d|ddk r|t|ddkr|t|tj |d|drt |tj |dnB|d}t t j||ftd dk}|d|}t |tj||d dk r|d dk rt|tj|d t |tj|d |d dk rHt|d dkrHt||d |d dk rt|d dkrt|tjt|tj|d t|tjd|krt|tj|dd|krt |tjt |dt|tjdS)z Write a noise covariance matrix.rirbrerrcNrhrar(rfrgrMrdrjrk)r"rrr$rrrrnr%rr&rrUrrrZravelrr'rrrZFIFFB_MNE_BAD_CHANNELSZFIFF_MNE_CH_NAME_LISTr#r(rrrx)ryrrbr[rrErErKrxs>         rxrr`cCs:t|dtf|td|ft|ts6t|dd}|S)Nz path-likezNoise covariance : %sFr`)r8r_r/rTrr)rrr0rErErK _ensure_cov"s   r)N)NN)rNrNNNrNrNNFTNN)TrjTN) TNNNrNrNNFrNN)N)TN)NNNrN)NNNFTrN)rrrrdTrrrrrrrrrrNNN)NNNN) NNNNFFFr3N)NNNNN)FN)r)ur{r itertoolsrmathrnumpyrUdefaultsrrrr Zio.writer Zio.projr r r rrrrriorrZio.pickrrrrrrrrZ io.constantsrZ io.meas_inforrZio.tagr Zio.treer!r"r#r$r%r&r'r(r)rXr*eventr+rr,rr-utilsr.r/r0r1r2r3r4r5r6r7r8r9r:r;r<r=r>rr?fixesr@rArBrCrDrPr^rr_rrrrrrr3rFr8rir[rZrUrWrr"rrrPrrrrrrxrrErErErKs   ((   $     L   &  $  - B >    >8I  : > 4 % o , r4