Module EEGToolkit.EEGStats.EEGStats
+Computes time-point-wise the mean and SEM values of EEG data stored in numpy ndarrays.
+
+
++Expand source code +
+"""
+Computes time-point-wise the mean and SEM values of EEG data stored in numpy ndarrays.
+"""
+
+import numpy as np
+from statsmodels.stats.weightstats import CompareMeans, DescrStatsW
+import matplotlib.pyplot as plt
+
+# for custom legend
+from matplotlib.lines import Line2D
+from matplotlib.patches import Patch
+
+# ----------------------------------------------------------------
+# Data Statistics
+# ----------------------------------------------------------------
+
+
+def mean(extracted_EEGData:np.ndarray) -> np.ndarray:
+ """
+ Computes the element-wise mean of an `m x n numpy ndarray`
+ with `m` repeated datasets and `n` entries per set along axis 1
+ (between repeated sets).
+
+ Parameters
+ ----------
+ extracted_EEGData : np.ndarray
+ An m x n numpy ndarray containing EEG data
+
+ Returns
+ -------
+ means : np.ndarray
+ An 1 x n numpy ndarray containing the element-wise means between all m repeated sets.
+ """
+ means = DescrStatsW( extracted_EEGData ).mean
+ return means
+
+def sem(extracted_EEGData:np.ndarray) -> np.ndarray:
+ """
+ Compute the standard error of the mean (SEM)
+ of an `m x n numpy ndarray` with `m` repeated
+ datasets and `n` entries per set along axis 1
+ (between repeated sets).
+
+ Parameters
+ ----------
+ extracted_EEGData : np.ndarray
+ An m x n numpy ndarray containing EEG data
+
+ Returns
+ -------
+ sems : np.ndarray
+ An 1 x n numpy ndarray containing the SEM between all m repeated sets.
+
+ """
+ sems = DescrStatsW( extracted_EEGData )
+ length = len( extracted_EEGData )
+ sems = sems.std / np.sqrt( length )
+ return sems
+
+
+
+def compare_signals(
+ extracted_EEGData_1 : np.ndarray,
+ extracted_EEGData_2 : np.ndarray
+ ) -> np.ndarray:
+ """
+ Compares two EEG signal datasets stored as identical `m x n` np ndarrays
+ with `m` repeated datasets and `n` entries per set, element-wise (along axis 1).
+ It performs T-Tests to check for siginificant differences between the two datasets
+ and returns the corresoponding p-values as a new array.
+
+ Parameters
+ ----------
+ extracted_EEGData_1 : np.ndarray
+ The first EEG signal dataset.
+
+ extracted_EEGData_2 : np.ndarray
+ The second EEG signal dataset.
+
+ Returns
+ -----
+ pvalues : np.ndarray
+ A 1 x n numpy ndarray of p-values for the
+ difference between Signals 1 and 2 at each position.
+ """
+
+ # compare the signal means
+ diff = CompareMeans.from_data(
+ extracted_EEGData_1,
+ extracted_EEGData_2
+ )
+
+ # perform t-test comparison
+ diff = diff.ttest_ind()
+
+ # get the p-values
+ pvalues = diff[1]
+ return pvalues
+
+# ----------------------------------------------------------------
+# Data Visualisation
+# ----------------------------------------------------------------
+
+# set up some default style settings
+
+# colorscheme
+signal_color = "gray"
+signal1_color = "xkcd:dark royal blue"
+signal2_color = "xkcd:watermelon"
+signif_shade_color = "xkcd:yellow tan"
+
+# opacities
+signal_alpha = 0.8
+sem_alpha = 0.2
+signif_shade_alpha = 0.5
+
+def plot_signal(
+ extracted_EEGData : np.ndarray,
+ sampling_frequency : float,
+ start_sec : float,
+ stop_sec : float,
+ x_scale = 10**3,
+ y_scale = 10**3,
+ baseline : np.ndarray = None,
+ significance_level:float = 0.05,
+ make_legend = False,
+ ax = None ) -> None:
+ """
+ Visualises a single EGG signal dataset from an `m x n numpy ndarray`
+ with `m` repeated datasets and `n` entries per set. It generates a solid
+ line for the time-point-wise mean and a shaded area of the corresponding SEM.
+
+ Parameters
+ ----------
+
+ extracted_EEGData : np.ndarray
+ An m x n numpy ndarray of repeated EEG data.
+
+ sampling_frequency : float
+ The frequency in which the EEG data was recorded in `Hertz`.
+
+ start_sec : float
+ The initial / low bound of the time-scale in seconds.
+
+ stop_sec : float
+ The final / upper bound of the time-scale in seconds.
+
+ x_scale : float
+ A scaling factor to adjust the data's x-value range.
+ E.g. `x_scale = 1000` to adjust the time-scale to milliseconds.
+
+ y_scale : float
+ A scaling factor for the data's y-value range.
+ E.g. `y_scale = 1000` to adjust the signal-scale to millivolts.
+
+ baseline : np.ndarray
+ An array of baseline data to compare the signal to using a pair-wise t-test.
+ This will introduce shaded boxes for regions that show significant differences
+ between the signal and the baseline.
+
+ significance_level : float
+ The significance threshold for which to accept a signal difference
+ as significant. Default is `0.05`.
+
+ make_legend : bool
+ Generates a legend for the plot.
+
+ Returns
+ -----
+ pvalues : np.ndarray
+ The p-value from pair-wise T-Test comparison between the
+ signal and the baseline (if provided) at each signal position (timepoint).
+ """
+
+ # generate a new figure if no ax is specified
+ if ax is None:
+ fig, ax = plt.subplots()
+ else:
+ fig = None
+
+ pvalues = _plot_(ax,
+ extracted_EEGData,
+ sampling_frequency,
+ start_sec,
+ stop_sec,
+ x_scale,
+ y_scale,
+ baseline,
+ make_legend,
+ significance_level = significance_level
+ )
+
+ # we show the figure only if no ax was provided
+ # and thus no "bigger" figure is assembled elsewhere...
+ if fig is not None:
+ plt.show()
+
+ return pvalues
+
+def difference_plot(extracted_EEGData_1:np.ndarray,
+ extracted_EEGData_2:np.ndarray,
+ sampling_frequency:float,
+ start_sec:float,
+ stop_sec:float,
+ significance_level:float = 0.05,
+ x_scale=10**3,
+ y_scale=10**3,
+ make_legend = False,
+ ax = None ) -> None:
+
+ """
+ Visualises the difference between two EEG signals and tests
+ time-point-wise the difference using T-Tests. Individual EEG
+ Signals are shown as mean-line with shaded SEM, and time-points
+ with significant differences are highlighted with overall-shading.
+
+ Parameters
+ ----------
+ extracted_EEGData_1 : np.ndarray
+ The first EEG signal dataset.
+
+ extracted_EEGData_2 : np.ndarray
+ The second EEG signal dataset.
+
+ sampling_frequency : float
+ The frequency in which the EEG data was recorded in `Hertz`.
+
+ significance_level : float
+ The significance threshold for which to accept a signal difference
+ as significant. Default is `0.05`.
+
+ start_sec : float
+ The initial / low bound of the time-scale in seconds.
+
+ stop_sec : float
+ The final / upper bound of the time-scale in seconds.
+
+ x_scale : float
+ A scaling factor to adjust the data's x-value range.
+ E.g. `x_scale = 1000` to adjust the time-scale to milliseconds.
+
+ y_scale : float
+ A scaling factor for the data's y-value range.
+ E.g. `y_scale = 1000` to adjust the signal-scale to millivolts.
+
+ make_legend : bool
+ Generates a legend for the plot.
+
+ Returns
+ -----
+ pvalues : np.ndarray
+ The p-value from pair-wise T-Test comparison between the
+ two signals at each signal position (timepoint).
+ """
+
+ # generate a new figure if no ax is given
+ if ax is None:
+ fig, ax = plt.subplots()
+ else:
+ fig = None
+
+ pvalues = _difference_plot_(ax,
+ extracted_EEGData_1,
+ extracted_EEGData_2,
+ significance_level,
+ sampling_frequency,
+ start_sec,
+ stop_sec,
+ x_scale,
+ y_scale,
+ make_legend
+ )
+
+ # we show the figure only if no ax was provided
+ # and thus no "bigger" figure is assembled elsewhere...
+ if fig is not None:
+ plt.show()
+
+ return pvalues
+
+def _plot_(
+ ax,
+ extracted_EEGData:np.ndarray,
+ sampling_frequency:float,
+ start_sec:float,
+ stop_sec:float,
+ x_scale=10**3,
+ y_scale=10**3,
+ baseline : np.ndarray = None,
+ make_legend = False,
+ **kwargs ) -> None:
+ """
+ Generates a mean signal line with shaded
+ SEM area from an m x n numpy ndarray.
+
+ Note
+ -----
+ This is the core for `plot_signal`
+ """
+ # compute mean and SEM for the signal
+ _mean = mean(extracted_EEGData)
+ _sem = sem(extracted_EEGData)
+
+ # now generate scaled xvalues
+ x_values = _scale_xboundries(sampling_frequency, start_sec, stop_sec, x_scale)
+
+ # now plot the signal's scaled mean line
+ signal = _mean * y_scale
+ ax.plot(x_values, signal, color = signal_color )
+
+ # and add the shading for SEM
+ lower = (_mean - _sem)
+ upper = (_mean + _sem)
+ lower *= y_scale
+ upper *= y_scale
+
+ ax.fill_between(
+ x_values,
+ lower,
+ upper,
+ color = signal_color,
+ edgecolor = None,
+ linewidth = 0,
+ alpha = sem_alpha
+
+ )
+
+ # if we have baseline data, we compare also to the baseline
+ # and shade siginificantly different regions...
+ pvalues = None
+ if baseline is not None:
+
+ pvalues = compare_signals( extracted_EEGData, baseline )
+
+ # generate the y-value boundries for the plot
+ # and scale the y-values to some user-defined range
+ # plus add a little padding
+ max_y = np.max( _mean )
+ min_y = np.min( _mean )
+ yvalues = _scale_yboundries( y_scale, max_y, min_y, pad = 1.2 )
+
+ # add the shaded fillings for significantly different
+ # timepoint areas
+ signif_level = kwargs.pop( "significance_level", 0.05 )
+ _shade_singificant_regions(
+ ax,
+ significance_level = signif_level,
+ pvalues = pvalues,
+ xvalues = x_values,
+ ylim = yvalues,
+ )
+
+
+
+ # and add a line for the start of the signal
+ ax.axvline( x=0, linewidth = 2, color = "black" )
+
+ # and some axes formatting...
+ ax.set_title("Average EEG Signal (Shaded area SEM)")
+ ax.set_ylabel("Signal\namplitude")
+ ax.set_xlabel("Time relative to event")
+
+ if make_legend:
+ handles = [
+ Line2D( [0], [0],
+ color = signal_color,
+ label = "Mean Signal"
+ ),
+ Patch(
+ facecolor = signal_color,
+ edgecolor = None, linewidth = 0,
+ alpha = sem_alpha,
+ label = "SEM"
+ )
+ ]
+ if baseline is not None:
+ handles.append(
+ Patch(
+ facecolor = signif_shade_color,
+ edgecolor = None, linewidth = 0,
+ alpha = signif_shade_alpha,
+ label = f"pvalue < {signif_level}"
+ )
+ )
+ # now add a custom legend
+ ax.legend( handles = handles,
+ bbox_to_anchor = (1, -0.5),
+ frameon = False
+ )
+ return pvalues
+
+def _difference_plot_(ax,
+ extracted_EEGData_1:np.ndarray,
+ extracted_EEGData_2:np.ndarray,
+ significance_level:float,
+ sampling_frequency:float,
+ start_sec:float,
+ stop_sec:float,
+ x_scale=10**3,
+ y_scale=10**3,
+ make_legend = False ) -> None:
+
+ """
+ Generates a plot of two EEG signals and their time-point-wise
+ difference using a T-Test.
+
+ Note
+ -----
+ This is the core of difference_plot
+ """
+
+ # compare the two EEG signals and generate a p-value
+ # from t-test position-wise comparisons...
+ pvalues = compare_signals(extracted_EEGData_1, extracted_EEGData_2)
+
+ # generate the mean lines for both signals
+ mean_1 = mean(extracted_EEGData_1)
+ mean_2 = mean(extracted_EEGData_2)
+
+ # generate the y-value boundries for the plot
+ max_y = np.max( [np.max(mean_1), np.max(mean_2)] )
+ min_y = np.min( [np.min(mean_1), np.min(mean_2)] )
+
+ # and scale the y-values to some user-defined range
+ # plus add a little padding
+ yvalues = _scale_yboundries( y_scale, max_y, min_y, pad = 1.2 )
+
+ # generate correspondingly scaled x-values
+ x_values = _scale_xboundries(sampling_frequency, start_sec, stop_sec, x_scale)
+
+ # add the shaded fillings for significantly different
+ # timepoint areas
+ _shade_singificant_regions(
+ ax,
+ significance_level = significance_level,
+ pvalues = pvalues,
+ xvalues = x_values,
+ ylim = yvalues,
+ )
+
+ # plot scaled signals 1 and 2
+ signal_1 = mean_1 * y_scale
+ signal_2 = mean_2 * y_scale
+
+ ax.plot(x_values, signal_1, color = signal1_color, alpha = signal_alpha )
+ ax.plot(x_values, signal_2, color = signal2_color, alpha = signal_alpha )
+
+ # plot a vertical line at the signal start
+ ax.axvline(x = 0, color = "black", linewidth = 2 )
+
+ # and some axes formatting...
+ ax.set_title("Average EEG Signal (Shaded area significant regions)")
+ ax.set_ylabel("Signal\namplitude")
+ ax.set_xlabel("Time relative to event")
+
+ if make_legend:
+ # now add a custom legend
+ ax.legend( handles = [
+ Line2D( [0], [0],
+ color = signal1_color,
+ label = "Mean horizontal Signal"
+ ),
+ Line2D( [0], [0],
+ color = signal2_color,
+ label = "Mean vertical Signal"
+ ),
+ Patch(
+ facecolor = signif_shade_color,
+ edgecolor = None, linewidth = 0,
+ alpha = signif_shade_alpha,
+ label = f"pvalue < {significance_level}"
+ )
+
+ ],
+ bbox_to_anchor = (1, -0.5),
+ frameon = False
+ )
+ return pvalues
+
+def _shade_singificant_regions(ax, significance_level : float , pvalues : np.ndarray , xvalues : np.ndarray , ylim : tuple ):
+ """
+ Shades the background of regions (x-value ranges) where corresponding p-values
+ are below a given significance level.
+
+ Parameters
+ ----------
+ ax : plt.axes.Axes
+ An Axes object to plot to.
+ significance_level : float
+ The significance level to use.
+ pvalues : np.ndarray
+ An array of pvalues for each x-value.
+ xvalues : np.ndarray
+ An array of x-values.
+ ylim : tuple
+ A tuple of minimal and maximal y-values to shade.
+ """
+ ax.fill_between(
+ xvalues,
+ ylim[0],
+ ylim[1],
+ # now fill areas where the pvalues are
+ # below our significance_level
+ where = pvalues < significance_level,
+ facecolor = signif_shade_color,
+ alpha = signif_shade_alpha
+ )
+
+def _scale_xboundries(sampling_frequency, start_sec, stop_sec, x_scale):
+ """
+ Scales minimal and maximal x values from starting and final timestep values
+ given some scale and sampling frequency.
+ """
+ x_values = np.arange(
+ int( start_sec * sampling_frequency ),
+ int( stop_sec * sampling_frequency ),
+ dtype=int
+ )
+ x_values = x_values / sampling_frequency * x_scale
+ return x_values
+
+def _scale_yboundries(y_scale, max_y, min_y, pad = 1 ):
+ """
+ Scales minimal and maximal y values from some data
+ by a given scale and adds additional padding as a scalar factor
+ (pad = 1 means no padding).
+ """
+ max_y *= y_scale * pad
+ min_y *= y_scale * pad
+ return min_y, max_yFunctions
+-
+
+def compare_signals(extracted_EEGData_1: numpy.ndarray, extracted_EEGData_2: numpy.ndarray) ‑> numpy.ndarray +
+-
++
Compares two EEG signal datasets stored as identical
+m x nnp ndarrays +withmrepeated datasets andnentries per set, element-wise (along axis 1). +It performs T-Tests to check for siginificant differences between the two datasets +and returns the corresoponding p-values as a new array.Parameters
+-
+
extracted_EEGData_1:np.ndarray
+- The first EEG signal dataset. +
extracted_EEGData_2:np.ndarray
+- The second EEG signal dataset. +
Returns
+-
+
pvalues:np.ndarray
+- A 1 x n numpy ndarray of p-values for the +difference between Signals 1 and 2 at each position. +
+++Expand source code +
+
+def compare_signals( + extracted_EEGData_1 : np.ndarray, + extracted_EEGData_2 : np.ndarray + ) -> np.ndarray: + """ + Compares two EEG signal datasets stored as identical `m x n` np ndarrays + with `m` repeated datasets and `n` entries per set, element-wise (along axis 1). + It performs T-Tests to check for siginificant differences between the two datasets + and returns the corresoponding p-values as a new array. + + Parameters + ---------- + extracted_EEGData_1 : np.ndarray + The first EEG signal dataset. + + extracted_EEGData_2 : np.ndarray + The second EEG signal dataset. + + Returns + ----- + pvalues : np.ndarray + A 1 x n numpy ndarray of p-values for the + difference between Signals 1 and 2 at each position. + """ + + # compare the signal means + diff = CompareMeans.from_data( + extracted_EEGData_1, + extracted_EEGData_2 + ) + + # perform t-test comparison + diff = diff.ttest_ind() + + # get the p-values + pvalues = diff[1] + return pvalues
+ +def difference_plot(extracted_EEGData_1: numpy.ndarray, extracted_EEGData_2: numpy.ndarray, sampling_frequency: float, start_sec: float, stop_sec: float, significance_level: float = 0.05, x_scale=1000, y_scale=1000, make_legend=False, ax=None) ‑> None +
+-
++
Visualises the difference between two EEG signals and tests +time-point-wise the difference using T-Tests. Individual EEG +Signals are shown as mean-line with shaded SEM, and time-points +with significant differences are highlighted with overall-shading.
+Parameters
+-
+
extracted_EEGData_1:np.ndarray
+- The first EEG signal dataset. +
extracted_EEGData_2:np.ndarray
+- The second EEG signal dataset. +
sampling_frequency:float
+- The frequency in which the EEG data was recorded in
Hertz.
+ significance_level:float
+- The significance threshold for which to accept a signal difference
+as significant. Default is
0.05.
+ start_sec:float +
+- The initial / low bound of the time-scale in seconds. +
stop_sec:float
+- The final / upper bound of the time-scale in seconds. +
x_scale:float
+- A scaling factor to adjust the data's x-value range.
+E.g.
x_scale = 1000to adjust the time-scale to milliseconds.
+ y_scale:float
+- A scaling factor for the data's y-value range.
+E.g.
y_scale = 1000to adjust the signal-scale to millivolts.
+ make_legend:bool
+- Generates a legend for the plot. +
Returns
+-
+
pvalues:np.ndarray +
+- The p-value from pair-wise T-Test comparison between the +two signals at each signal position (timepoint). +
+++Expand source code +
+
+def difference_plot(extracted_EEGData_1:np.ndarray, + extracted_EEGData_2:np.ndarray, + sampling_frequency:float, + start_sec:float, + stop_sec:float, + significance_level:float = 0.05, + x_scale=10**3, + y_scale=10**3, + make_legend = False, + ax = None ) -> None: + + """ + Visualises the difference between two EEG signals and tests + time-point-wise the difference using T-Tests. Individual EEG + Signals are shown as mean-line with shaded SEM, and time-points + with significant differences are highlighted with overall-shading. + + Parameters + ---------- + extracted_EEGData_1 : np.ndarray + The first EEG signal dataset. + + extracted_EEGData_2 : np.ndarray + The second EEG signal dataset. + + sampling_frequency : float + The frequency in which the EEG data was recorded in `Hertz`. + + significance_level : float + The significance threshold for which to accept a signal difference + as significant. Default is `0.05`. + + start_sec : float + The initial / low bound of the time-scale in seconds. + + stop_sec : float + The final / upper bound of the time-scale in seconds. + + x_scale : float + A scaling factor to adjust the data's x-value range. + E.g. `x_scale = 1000` to adjust the time-scale to milliseconds. + + y_scale : float + A scaling factor for the data's y-value range. + E.g. `y_scale = 1000` to adjust the signal-scale to millivolts. + + make_legend : bool + Generates a legend for the plot. + + Returns + ----- + pvalues : np.ndarray + The p-value from pair-wise T-Test comparison between the + two signals at each signal position (timepoint). + """ + + # generate a new figure if no ax is given + if ax is None: + fig, ax = plt.subplots() + else: + fig = None + + pvalues = _difference_plot_(ax, + extracted_EEGData_1, + extracted_EEGData_2, + significance_level, + sampling_frequency, + start_sec, + stop_sec, + x_scale, + y_scale, + make_legend + ) + + # we show the figure only if no ax was provided + # and thus no "bigger" figure is assembled elsewhere... + if fig is not None: + plt.show() + + return pvalues
+ +def mean(extracted_EEGData: numpy.ndarray) ‑> numpy.ndarray +
+-
++
Computes the element-wise mean of an
+m x n numpy ndarray+withmrepeated datasets andnentries per set along axis 1 +(between repeated sets).Parameters
+-
+
extracted_EEGData:np.ndarray
+- An m x n numpy ndarray containing EEG data +
Returns
+-
+
means:np.ndarray
+- An 1 x n numpy ndarray containing the element-wise means between all m repeated sets. +
+++Expand source code +
+
+def mean(extracted_EEGData:np.ndarray) -> np.ndarray: + """ + Computes the element-wise mean of an `m x n numpy ndarray` + with `m` repeated datasets and `n` entries per set along axis 1 + (between repeated sets). + + Parameters + ---------- + extracted_EEGData : np.ndarray + An m x n numpy ndarray containing EEG data + + Returns + ------- + means : np.ndarray + An 1 x n numpy ndarray containing the element-wise means between all m repeated sets. + """ + means = DescrStatsW( extracted_EEGData ).mean + return means
+ +def plot_signal(extracted_EEGData: numpy.ndarray, sampling_frequency: float, start_sec: float, stop_sec: float, x_scale=1000, y_scale=1000, baseline: numpy.ndarray = None, significance_level: float = 0.05, make_legend=False, ax=None) ‑> None +
+-
++
Visualises a single EGG signal dataset from an
+m x n numpy ndarray+withmrepeated datasets andnentries per set. It generates a solid +line for the time-point-wise mean and a shaded area of the corresponding SEM.Parameters
+-
+
extracted_EEGData:np.ndarray
+- An m x n numpy ndarray of repeated EEG data. +
sampling_frequency:float
+- The frequency in which the EEG data was recorded in
Hertz.
+ start_sec:float +
+- The initial / low bound of the time-scale in seconds. +
stop_sec:float
+- The final / upper bound of the time-scale in seconds. +
x_scale:float
+- A scaling factor to adjust the data's x-value range.
+E.g.
x_scale = 1000to adjust the time-scale to milliseconds.
+ y_scale:float
+- A scaling factor for the data's y-value range.
+E.g.
y_scale = 1000to adjust the signal-scale to millivolts.
+ baseline:np.ndarray
+- An array of baseline data to compare the signal to using a pair-wise t-test. +This will introduce shaded boxes for regions that show significant differences +between the signal and the baseline. +
significance_level:float
+- The significance threshold for which to accept a signal difference
+as significant. Default is
0.05.
+ make_legend:bool
+- Generates a legend for the plot. +
Returns
+-
+
pvalues:np.ndarray +
+- The p-value from pair-wise T-Test comparison between the +signal and the baseline (if provided) at each signal position (timepoint). +
+++Expand source code +
+
+def plot_signal( + extracted_EEGData : np.ndarray, + sampling_frequency : float, + start_sec : float, + stop_sec : float, + x_scale = 10**3, + y_scale = 10**3, + baseline : np.ndarray = None, + significance_level:float = 0.05, + make_legend = False, + ax = None ) -> None: + """ + Visualises a single EGG signal dataset from an `m x n numpy ndarray` + with `m` repeated datasets and `n` entries per set. It generates a solid + line for the time-point-wise mean and a shaded area of the corresponding SEM. + + Parameters + ---------- + + extracted_EEGData : np.ndarray + An m x n numpy ndarray of repeated EEG data. + + sampling_frequency : float + The frequency in which the EEG data was recorded in `Hertz`. + + start_sec : float + The initial / low bound of the time-scale in seconds. + + stop_sec : float + The final / upper bound of the time-scale in seconds. + + x_scale : float + A scaling factor to adjust the data's x-value range. + E.g. `x_scale = 1000` to adjust the time-scale to milliseconds. + + y_scale : float + A scaling factor for the data's y-value range. + E.g. `y_scale = 1000` to adjust the signal-scale to millivolts. + + baseline : np.ndarray + An array of baseline data to compare the signal to using a pair-wise t-test. + This will introduce shaded boxes for regions that show significant differences + between the signal and the baseline. + + significance_level : float + The significance threshold for which to accept a signal difference + as significant. Default is `0.05`. + + make_legend : bool + Generates a legend for the plot. + + Returns + ----- + pvalues : np.ndarray + The p-value from pair-wise T-Test comparison between the + signal and the baseline (if provided) at each signal position (timepoint). + """ + + # generate a new figure if no ax is specified + if ax is None: + fig, ax = plt.subplots() + else: + fig = None + + pvalues = _plot_(ax, + extracted_EEGData, + sampling_frequency, + start_sec, + stop_sec, + x_scale, + y_scale, + baseline, + make_legend, + significance_level = significance_level + ) + + # we show the figure only if no ax was provided + # and thus no "bigger" figure is assembled elsewhere... + if fig is not None: + plt.show() + + return pvalues
+ +def sem(extracted_EEGData: numpy.ndarray) ‑> numpy.ndarray +
+-
++
Compute the standard error of the mean (SEM) +of an
+m x n numpy ndarraywithmrepeated +datasets andnentries per set along axis 1 +(between repeated sets).Parameters
+-
+
extracted_EEGData:np.ndarray
+- An m x n numpy ndarray containing EEG data +
Returns
+-
+
sems:np.ndarray
+- An 1 x n numpy ndarray containing the SEM between all m repeated sets. +
+++Expand source code +
+
+def sem(extracted_EEGData:np.ndarray) -> np.ndarray: + """ + Compute the standard error of the mean (SEM) + of an `m x n numpy ndarray` with `m` repeated + datasets and `n` entries per set along axis 1 + (between repeated sets). + + Parameters + ---------- + extracted_EEGData : np.ndarray + An m x n numpy ndarray containing EEG data + + Returns + ------- + sems : np.ndarray + An 1 x n numpy ndarray containing the SEM between all m repeated sets. + + """ + sems = DescrStatsW( extracted_EEGData ) + length = len( extracted_EEGData ) + sems = sems.std / np.sqrt( length ) + return sems
+