emgdecompy.contrast

Module Contents

Functions

skew(w, der=False)	Applies contrast function (if der=False) or
log_cosh(w, der=False)	Applies contrast function (if der=False) or
exp_sq(x, der=False)	Applies contrast function (if der=False) or
apply_contrast(w, fun=skew, der=False)	Takes first derivitive and applies contrast function to w

emgdecompy.contrast.skew(w, der=False)

Applies contrast function (if der=False) or first derivative of contrast function (if der=True) to w. skew = w^3 / 3

Parameters

• w (np.array) – Array to apply contrast function to.

• der (boolean) – Whether to apply derivative (or base version).

Returns

Array with contrast function applied, same shape as w.

Return type

np.array

Examples

>>> w = np.array([1, 2, 3, 800])
>>> skew(w, der=True)
array([1, 4, 9, 640000])

emgdecompy.contrast.log_cosh(w, der=False)

Applies contrast function (if der=False) or first derivative of contrast function (if der=True) to each element of w. function = log(cosh(w))

Parameters

• w (np.array) – Array to apply contrast function to.

• der (boolean) – Whether to apply derivative (or base version).

Returns

Array with contrast function applied, same shape as w.

Return type

np.array

Examples

>>> w = np.array([1, 2, 3, 800])
>>> log_cosh(w)
array([4.33780830e-01, 1.32500275e+00, 2.30932850e+00, 7.99300000e+02])

emgdecompy.contrast.exp_sq(x, der=False)

Applies contrast function (if der=False) or first derivative of contrast function (if der=True) to w. exp_sq = exp((-x^2/2))

Parameters

• w (np.array) – Array to apply contrast function to.

• der (boolean) – Whether to apply derivative (or base version).

Returns

Array with contrast function applied, same shape as w.

Return type

np.array

Examples

>>> w = np.array([1, 2, 3, 800])
>>> exp_sq(w, der=False)
array([0.60653066, 0.13533528, 0.011109, 0.])

emgdecompy.contrast.apply_contrast(w, fun=skew, der=False)

Takes first derivitive and applies contrast function to w for Step 2a of fixed point algorithm. Options include functions mentioned in Negro et al. (2016).

Parameters

• fun (str) – Name of contrast function to use.

• w (numpy.ndarray) – Matrix to apply contrast function to.

Returns

Matrix with contrast function applied.

Return type

numpy.ndarray

Examples

>>> w = np.array([1, 2, 3])
>>> fun = skew
>>> apply_contrast(w, fun, True)
array([1, 4, 9])

>>> w = np.array([0.01, 0.1, 1, 10, 100, 1000])
>>> fun = log_cosh
>>> apply_contrast(w, fun)
array([4.99991667e-05, 4.99168882e-03, 4.33780830e-01, 9.30685282e+00,
9.93068528e+01, 9.99300000e+02])