emgdecompy.preprocessing

Module Contents

Functions

flatten_signal(raw)	Takes the raw EMG signal array, flattens it, and removes empty channels with no data.
butter_bandpass_filter(data, lowcut=10, highcut=900, fs=2048, order=6)	Filters input data using a Butterworth band-pass filter.
extend_input_by_R(x, R)	Takes a one-dimensional array and extends it using past observations.
extend_all_channels(x_mat, R)	Takes an array with dimensions M by K,
center_matrix(x)	Subtract mean of each row.
whiten(x)	Whiten the input matrix through zero component analysis.

emgdecompy.preprocessing.flatten_signal(raw)

Takes the raw EMG signal array, flattens it, and removes empty channels with no data.

Parameters

raw (numpy.ndarray) – Raw EMG signal array.

Returns

Flattened EMG signal array, with empty channels removed.

Return type

numpy.ndarray

emgdecompy.preprocessing.butter_bandpass_filter(data, lowcut=10, highcut=900, fs=2048, order=6)

Filters input data using a Butterworth band-pass filter.

Parameters

• data (numpy.ndarray) – 1D array containing data to be filtered.

• lowcut (float) – Lower range of band-pass filter.

• highcut (float) – Upper range of band-pass filter.

• fs (float) – Sampling frequency in Hz.

• order (int) – Order of filter.

Returns

Filtered data.

Return type

numpy.ndarray

Examples

>>> butter_bandpass_filter(data, 10, 900, 2048, order=6)

emgdecompy.preprocessing.extend_input_by_R(x, R)

Takes a one-dimensional array and extends it using past observations.

Parameters

• x (numpy.ndarray) – 1D array to be extended.

• R (int) – How far to extend x.

Returns

len(x) by R+1 extended array.

Return type

numpy.ndarray

Examples

>>> R = 5
>>> x = np.array([1, 2, 3])
>>> extend_input_by_R(x, R)
array([[1., 2., 3.],
       [0., 1., 2.],
       [0., 0., 1.],
       [0., 0., 0.],
       [0., 0., 0.],
       [0., 0., 0.]])

emgdecompy.preprocessing.extend_all_channels(x_mat, R)

Takes an array with dimensions M by K, where M represents number of channels and K represents observations, and “extends” it to return an array of shape M * (R+1) by K.

Parameters

• x_mat (numpy.ndarray) – 2D array to be extended.

• R (int) – How far to extend x.

Returns

M(R+1) x K extended array.

Return type

numpy.ndarray

Examples

>>> R = 3
>>> x_mat = np.array([[1, 2, 3, 4,], [5, 6, 7, 8,]])
>>> extend_all_channels(x_mat, R)
array([[1., 2., 3., 4.],
       [0., 1., 2., 3.],
       [0., 0., 1., 2.],
       [0., 0., 0., 1.],
       [5., 6., 7., 8.],
       [0., 5., 6., 7.],
       [0., 0., 5., 6.],
       [0., 0., 0., 5.]])

emgdecompy.preprocessing.center_matrix(x)

Subtract mean of each row. Results in the data being centered around x=0.

Parameters

x (numpy.ndarray) – Matrix of arrays to be centered.

Returns

Centered matrix array.

Return type

numpy.ndarray

Examples

>>> x = np.array([[1, 2, 3], [4, 6, 8]])
>>> center_matrix(x)
array([[-1.,  0.,  1.],
       [-2.,  0.,  2.]])

emgdecompy.preprocessing.whiten(x)

Whiten the input matrix through zero component analysis.

Parameters

x (numpy.ndarray) – Centred 2D array to be whitened.

Returns

Whitened array.

Return type

numpy.ndarray

Examples

>>> x = np.array([[1, 2, 3, 4],  # Feature-1
                  [5, 6, 7, 8]]) # Feature-2
>>> whiten(x)
array([[-1.34217726e+08, -1.34217725e+08, -1.34217725e+08, -1.34217724e+08],
       [ 1.34217730e+08,  1.34217731e+08,  1.34217731e+08, 1.34217732e+08]])