mne.decoding.CSP¶

class mne.decoding.CSP(n_components=4, reg=None, log=None, cov_est=’concat’, transform_into=’average_power’)[source]¶

M/EEG signal decomposition using the Common Spatial Patterns (CSP).

This object can be used as a supervised decomposition to estimate spatial filters for feature extraction in a 2 class decoding problem. CSP in the context of EEG was first described in [1]; a comprehensive tutorial on CSP can be found in [2]. Multiclass solving is implemented from [3].

Parameters:

n_components : int, defaults to 4 The number of components to decompose M/EEG signals. This number should be set by cross-validation. reg : float | str | None, defaults to None if not None, allow regularization for covariance estimation if float, shrinkage covariance is used (0 <= shrinkage <= 1). if str, optimal shrinkage using Ledoit-Wolf Shrinkage (‘ledoit_wolf’) or Oracle Approximating Shrinkage (‘oas’). log : None | bool, defaults to None If transform_into == ‘average_power’ and log is None or True, then applies a log transform to standardize the features, else the features are z-scored. If transform_into == ‘csp_space’, then log must be None. cov_est : ‘concat’ | ‘epoch’, defaults to ‘concat’ If ‘concat’, covariance matrices are estimated on concatenated epochs for each class. If ‘epoch’, covariance matrices are estimated on each epoch separately and then averaged over each class. transform_into : {‘average_power’, ‘csp_space’} If ‘average_power’ then self.transform will return the average power of each spatial filter. If ‘csp_space’ self.transform will return the data in CSP space. Defaults to ‘average_power’.

References

[1] Zoltan J. Koles, Michael S. Lazar, Steven Z. Zhou. Spatial Patterns

Underlying Population Differences in the Background EEG. Brain Topography 2(4), 275-284, 1990.

[2] Benjamin Blankertz, Ryota Tomioka, Steven Lemm, Motoaki Kawanabe,

Klaus-Robert Müller. Optimizing Spatial Filters for Robust EEG Single-Trial Analysis. IEEE Signal Processing Magazine 25(1), 41-56, 2008.

[3] Grosse-Wentrup, Moritz, and Martin Buss. Multiclass common spatial

patterns and information theoretic feature extraction. IEEE Transactions on Biomedical Engineering, Vol 55, no. 8, 2008.

Attributes

filters_	(ndarray, shape (n_channels, n_channels)) If fit, the CSP components used to decompose the data, else None.
patterns_	(ndarray, shape (n_channels, n_channels)) If fit, the CSP patterns used to restore M/EEG signals, else None.
mean_	(ndarray, shape (n_components,)) If fit, the mean squared power for each component.
std_	(ndarray, shape (n_components,)) If fit, the std squared power for each component.

Methods

__hash__() <==> hash(x)
fit(X, y)	Estimate the CSP decomposition on epochs.
fit_transform(X[, y])	Fit to data, then transform it.
get_params([deep])	Get parameters for this estimator.
plot_filters(info[, components, ch_type, …])	Plot topographic filters of CSP components.
plot_patterns(info[, components, ch_type, …])	Plot topographic patterns of CSP components.
set_params(**params)	Set the parameters of this estimator.
transform(X)	Estimate epochs sources given the CSP filters.

__hash__() <==> hash(x)¶

fit(X, y)[source]¶

Estimate the CSP decomposition on epochs.

Parameters:	X : ndarray, shape (n_epochs, n_channels, n_times) The data on which to estimate the CSP. y : array, shape (n_epochs,) The class for each epoch.
Returns:	self : instance of CSP Returns the modified instance.

fit_transform(X, y=None, **fit_params)[source]¶

Fit to data, then transform it.

Fits transformer to X and y with optional parameters fit_params and returns a transformed version of X.

Parameters:	X : numpy array of shape [n_samples, n_features] Training set. y : numpy array of shape [n_samples] Target values.
Returns:	X_new : numpy array of shape [n_samples, n_features_new] Transformed array.

get_params(deep=True)[source]¶

Get parameters for this estimator.

Parameters:	deep : boolean, optional If True, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:	params : mapping of string to any Parameter names mapped to their values.

plot_filters(info, components=None, ch_type=None, layout=None, vmin=None, vmax=None, cmap=’RdBu_r’, sensors=True, colorbar=True, scale=None, scale_time=1, unit=None, res=64, size=1, cbar_fmt=’%3.1f’, name_format=’CSP%01d’, proj=False, show=True, show_names=False, title=None, mask=None, mask_params=None, outlines=’head’, contours=6, image_interp=’bilinear’, average=None, head_pos=None)[source]¶

Plot topographic filters of CSP components.

The CSP filters are used to extract discriminant neural sources from the measured data (a.k.a. the backward model).

Parameters:

info : instance of Info Info dictionary of the epochs used to fit CSP. If not possible, consider using create_info. components : float | array of floats | None. The CSP patterns to plot. If None, n_components will be shown. ch_type : ‘mag’ | ‘grad’ | ‘planar1’ | ‘planar2’ | ‘eeg’ | None The channel type to plot. For ‘grad’, the gradiometers are collected in pairs and the RMS for each pair is plotted. If None, then first available channel type from order given above is used. Defaults to None. layout : None | Layout Layout instance specifying sensor positions (does not need to be specified for Neuromag data). If possible, the correct layout file is inferred from the data; if no appropriate layout file was found the layout is automatically generated from the sensor locations. vmin : float | callable The value specfying the lower bound of the color range. If None, and vmax is None, -vmax is used. Else np.min(data). If callable, the output equals vmin(data). vmax : float | callable The value specfying the upper bound of the color range. If None, the maximum absolute value is used. If vmin is None, but vmax is not, defaults to np.min(data). If callable, the output equals vmax(data). cmap : matplotlib colormap | (colormap, bool) | ‘interactive’ | None Colormap to use. If tuple, the first value indicates the colormap to use and the second value is a boolean defining interactivity. In interactive mode the colors are adjustable by clicking and dragging the colorbar with left and right mouse button. Left mouse button moves the scale up and down and right mouse button adjusts the range. Hitting space bar resets the range. Up and down arrows can be used to change the colormap. If None, ‘Reds’ is used for all positive data, otherwise defaults to ‘RdBu_r’. If ‘interactive’, translates to (None, True). Defaults to ‘RdBu_r’. Warning Interactive mode works smoothly only for a small amount of topomaps. sensors : bool | str Add markers for sensor locations to the plot. Accepts matplotlib plot format string (e.g., ‘r+’ for red plusses). If True, a circle will be used (via .add_artist). Defaults to True. colorbar : bool Plot a colorbar. scale : dict | float | None Scale the data for plotting. If None, defaults to 1e6 for eeg, 1e13 for grad and 1e15 for mag. scale_time : float | None Scale the time labels. Defaults to 1. unit : dict | str | None The unit of the channel type used for colorbar label. If scale is None the unit is automatically determined. res : int The resolution of the topomap image (n pixels along each side). size : float Side length per topomap in inches. cbar_fmt : str String format for colorbar values. name_format : str String format for topomap values. Defaults to “CSP%01d” proj : bool | ‘interactive’ If true SSP projections are applied before display. If ‘interactive’, a check box for reversible selection of SSP projection vectors will be show. show : bool Show figure if True. show_names : bool | callable If True, show channel names on top of the map. If a callable is passed, channel names will be formatted using the callable; e.g., to delete the prefix ‘MEG ‘ from all channel names, pass the function lambda x: x.replace(‘MEG ‘, ”). If mask is not None, only significant sensors will be shown. title : str | None Title. If None (default), no title is displayed. mask : ndarray of bool, shape (n_channels, n_times) | None The channels to be marked as significant at a given time point. Indices set to True will be considered. Defaults to None. mask_params : dict | None Additional plotting parameters for plotting significant sensors. Default (None) equals: dict(marker='o', markerfacecolor='w', markeredgecolor='k', linewidth=0, markersize=4) outlines : ‘head’ | ‘skirt’ | dict | None The outlines to be drawn. If ‘head’, the default head scheme will be drawn. If ‘skirt’ the head scheme will be drawn, but sensors are allowed to be plotted outside of the head circle. If dict, each key refers to a tuple of x and y positions, the values in ‘mask_pos’ will serve as image mask, and the ‘autoshrink’ (bool) field will trigger automated shrinking of the positions due to points outside the outline. Alternatively, a matplotlib patch object can be passed for advanced masking options, either directly or as a function that returns patches (required for multi-axis plots). If None, nothing will be drawn. Defaults to ‘head’. contours : int | False | None The number of contour lines to draw. If 0, no contours will be drawn. image_interp : str The image interpolation to be used. All matplotlib options are accepted. average : float | None The time window around a given time to be used for averaging (seconds). For example, 0.01 would translate into window that starts 5 ms before and ends 5 ms after a given time point. Defaults to None, which means no averaging. head_pos : dict | None If None (default), the sensors are positioned such that they span the head circle. If dict, can have entries ‘center’ (tuple) and ‘scale’ (tuple) for what the center and scale of the head should be relative to the electrode locations.

Returns:

fig : instance of matplotlib.figure.Figure The figure.

plot_patterns(info, components=None, ch_type=None, layout=None, vmin=None, vmax=None, cmap=’RdBu_r’, sensors=True, colorbar=True, scale=None, scale_time=1, unit=None, res=64, size=1, cbar_fmt=’%3.1f’, name_format=’CSP%01d’, proj=False, show=True, show_names=False, title=None, mask=None, mask_params=None, outlines=’head’, contours=6, image_interp=’bilinear’, average=None, head_pos=None)[source]¶

Plot topographic patterns of CSP components.

The CSP patterns explain how the measured data was generated from the neural sources (a.k.a. the forward model).

Parameters:

info : instance of Info Info dictionary of the epochs used to fit CSP. If not possible, consider using create_info. components : float | array of floats | None. The CSP patterns to plot. If None, n_components will be shown. ch_type : ‘mag’ | ‘grad’ | ‘planar1’ | ‘planar2’ | ‘eeg’ | None The channel type to plot. For ‘grad’, the gradiometers are collected in pairs and the RMS for each pair is plotted. If None, then first available channel type from order given above is used. Defaults to None. layout : None | Layout Layout instance specifying sensor positions (does not need to be specified for Neuromag data). If possible, the correct layout file is inferred from the data; if no appropriate layout file was found the layout is automatically generated from the sensor locations. vmin : float | callable The value specfying the lower bound of the color range. If None, and vmax is None, -vmax is used. Else np.min(data). If callable, the output equals vmin(data). vmax : float | callable The value specfying the upper bound of the color range. If None, the maximum absolute value is used. If vmin is None, but vmax is not, defaults to np.min(data). If callable, the output equals vmax(data). cmap : matplotlib colormap | (colormap, bool) | ‘interactive’ | None Colormap to use. If tuple, the first value indicates the colormap to use and the second value is a boolean defining interactivity. In interactive mode the colors are adjustable by clicking and dragging the colorbar with left and right mouse button. Left mouse button moves the scale up and down and right mouse button adjusts the range. Hitting space bar resets the range. Up and down arrows can be used to change the colormap. If None, ‘Reds’ is used for all positive data, otherwise defaults to ‘RdBu_r’. If ‘interactive’, translates to (None, True). Defaults to ‘RdBu_r’. Warning Interactive mode works smoothly only for a small amount of topomaps. sensors : bool | str Add markers for sensor locations to the plot. Accepts matplotlib plot format string (e.g., ‘r+’ for red plusses). If True, a circle will be used (via .add_artist). Defaults to True. colorbar : bool Plot a colorbar. scale : dict | float | None Scale the data for plotting. If None, defaults to 1e6 for eeg, 1e13 for grad and 1e15 for mag. scale_time : float | None Scale the time labels. Defaults to 1. unit : dict | str | None The unit of the channel type used for colorbar label. If scale is None the unit is automatically determined. res : int The resolution of the topomap image (n pixels along each side). size : float Side length per topomap in inches. cbar_fmt : str String format for colorbar values. name_format : str String format for topomap values. Defaults to “CSP%01d” proj : bool | ‘interactive’ If true SSP projections are applied before display. If ‘interactive’, a check box for reversible selection of SSP projection vectors will be show. show : bool Show figure if True. show_names : bool | callable If True, show channel names on top of the map. If a callable is passed, channel names will be formatted using the callable; e.g., to delete the prefix ‘MEG ‘ from all channel names, pass the function lambda x: x.replace(‘MEG ‘, ”). If mask is not None, only significant sensors will be shown. title : str | None Title. If None (default), no title is displayed. mask : ndarray of bool, shape (n_channels, n_times) | None The channels to be marked as significant at a given time point. Indices set to True will be considered. Defaults to None. mask_params : dict | None Additional plotting parameters for plotting significant sensors. Default (None) equals: dict(marker='o', markerfacecolor='w', markeredgecolor='k', linewidth=0, markersize=4) outlines : ‘head’ | ‘skirt’ | dict | None The outlines to be drawn. If ‘head’, the default head scheme will be drawn. If ‘skirt’ the head scheme will be drawn, but sensors are allowed to be plotted outside of the head circle. If dict, each key refers to a tuple of x and y positions, the values in ‘mask_pos’ will serve as image mask, and the ‘autoshrink’ (bool) field will trigger automated shrinking of the positions due to points outside the outline. Alternatively, a matplotlib patch object can be passed for advanced masking options, either directly or as a function that returns patches (required for multi-axis plots). If None, nothing will be drawn. Defaults to ‘head’. contours : int | False | None The number of contour lines to draw. If 0, no contours will be drawn. image_interp : str The image interpolation to be used. All matplotlib options are accepted. average : float | None The time window around a given time to be used for averaging (seconds). For example, 0.01 would translate into window that starts 5 ms before and ends 5 ms after a given time point. Defaults to None, which means no averaging. head_pos : dict | None If None (default), the sensors are positioned such that they span the head circle. If dict, can have entries ‘center’ (tuple) and ‘scale’ (tuple) for what the center and scale of the head should be relative to the electrode locations.

Returns:

fig : instance of matplotlib.figure.Figure The figure.

set_params(**params)[source]¶

Set the parameters of this estimator. The method works on simple estimators as well as on nested objects (such as pipelines). The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object. Returns ——- self

transform(X)[source]¶

Estimate epochs sources given the CSP filters.

Parameters:	X : array, shape (n_epochs, n_channels, n_times) The data.
Returns:	X : ndarray If self.transform_into == ‘average_power’ then returns the power of CSP features averaged over time and shape (n_epochs, n_sources) If self.transform_into == ‘csp_space’ then returns the data in CSP space and shape is (n_epochs, n_sources, n_times)