Merge branch 'decomposition' into test-separation-feat

src/emgdecompy/contrast.py

@@ -1,73 +1,78 @@
 import numpy as np
+import warnings
 
 
-def skew(x, der=False):
+def skew(w, der=False):
     """
     Applies contrast function (if der=False) or
     first derivative of contrast function (if der=True)
     to w.
-    skew = x^3 / 3
+    skew = w^3 / 3
 
     Parameters
     ----------
-        x: float
-            Number to apply contrast function to.
+        w: np.array
+            Array to apply contrast function to.
         der: boolean
             Whether to apply derivative (or base version).
 
     Returns
     -------
-        float
-            Float with contrast function applied.
+        np.array
+            Array with contrast function applied, same shape as w.
 
     Examples
     --------
-        >>> x = 4
-        >>> skew(x, der=True)
-        16
+        >>> w = np.array([1, 2, 3, 800])
+        >>> skew(w, der=True)
+        array([1, 4, 9, 640000])
     """
 
-    # first derivative of x^3/3 = x^2
+    # first derivitive of x^3/3 = x^2
     if der == True:
-        rtn = x ** 2
+        rtn = w ** 2
     else:
-        rtn = (x ** 3) / 3
+        rtn = (w ** 3) / 3
 
     return rtn
 
 
-def log_cosh(x, der=False):
+def log_cosh(w, der=False):
     """
     Applies contrast function (if der=False) or
     first derivative of contrast function (if der=True)
-    to w.
-    function = log(cosh(x))
+    to each element of w.
+    function = log(cosh(w))
+
     Parameters
     ----------
-        x: float
-            Number to apply contrast function to.
+        w: np.array
+            Array to apply contrast function to.
         der: boolean
             Whether to apply derivative (or base version).
+
     Returns
     -------
-        float
-            Float with contrast function applied.
+        np.array
+            Array with contrast function applied, same shape as w.
+
     Examples
     --------
-        >>> x = 4
-        >>> log_cosh(x)
-        3.3071882258129506
+        >>> w = np.array([1, 2, 3, 800])
+        >>> log_cosh(w)
+        array([4.33780830e-01, 1.32500275e+00, 2.30932850e+00, 7.99300000e+02])
     """
 
-    # first derivative of log(cosh(x)) = tanh(x)
+    # First derivitive of log(cosh(x)) = tanh(x)
     if der == True:
-        rtn = np.tanh(x)
+        rtn = np.tanh(w)
     else:
-        x = abs(x)
-        if x > 710:  # cosh(x) breaks for abs(x) > 710
-            rtn = x - 0.7
-        else:
-            rtn = np.log(np.cosh(x))
+        warnings.filterwarnings(
+            "ignore"
+        )  # To avoid warning from np.cosh(w) for values over 710
+        w = abs(w)
+        rtn = np.where(w > 710, w - 0.7, np.log(np.cosh(w)))
+        warnings.resetwarnings()
 
     return rtn
 
@@ -77,28 +82,28 @@ def exp_sq(x, der=False):
     Applies contrast function (if der=False) or
     first derivative of contrast function (if der=True)
     to w.
-    function = exp((-x^2/2))
+    exp_sq = exp((-x^2/2))
 
     Parameters
     ----------
-        x: float
-            Number to apply contrast function to.
+        w: np.array
+            Array to apply contrast function to.
         der: boolean
             Whether to apply derivative (or base version).
 
     Returns
     -------
-        float
-            Float with contrast function applied.
+        np.array
+            Array with contrast function applied, same shape as w.
 
     Examples
     --------
-        >>> x = 4
-        >>> exp_sq(4, der=True)
-        -0.0013418505116100474
+        >>> w = np.array([1, 2, 3, 800])
+        >>> exp_sq(w, der=False)
+        array([0.60653066, 0.13533528, 0.011109, 0.])
     """
 
-    # first derivative of exp((-x^2/2)) = -e^(-x^2/2) x
+    # first derivitive of exp((-x^2/2)) = -e^(-x^2/2) x
     pwr_x = -(x ** 2) / 2
     if der == True:
         rtn = -(np.exp(pwr_x) * x)
@@ -130,8 +135,15 @@ def apply_contrast(w, fun=skew, der=False):
     --------
         >>> w = np.array([1, 2, 3])
         >>> fun = skew
-        >>> apply_contrast(w, fun)
+        >>> 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])
+
     """
 
     rtn = fun(w, der)

tests/test_decomposition.py

@@ -88,4 +88,26 @@ def test_separation():
         if n > max_iter:
             break
 
-    assert (w_curr == emg.decomposition.separation(z, B, Tolx, emg.contrast.skew, max_iter)).all(), "Separation vector incorrectly calculated."
+    assert (w_curr == emg.decomposition.separation(z, B, Tolx, emg.contrast.skew, max_iter)).all(), "Separation vector incorrectly calculated."
+
+def test_orthogonalize():
+    """
+    Run unit tests on orthogonalize() function from EMGdecomPy.
+    """
+    for i in range(0, 10):
+        x = np.random.randint(1, 100)
+        y = np.random.randint(1, 100)
+        z = np.random.randint(1, 100) 
+
+        w = np.random.randn(x, y) * 1000 
+        b = np.random.randn(x, z) * 1000
+
+        assert b.T.shape[1] == w.shape[0], "Dimensions of input arrays are not compatible."
+
+        # orthogonalize by hand 
+        ortho = w - b @ (b.T @ w)
+
+        fx = emg.decomposition.orthogonalize(w, b)
+
+        assert np.array_equal(ortho, fx), "Manually calculated array not equivalent to emg.orthogonalize()"
+        assert fx.shape == w.shape, "The output shape of w is incorrect."