Merge branch 'master' into fix_smoothing

demos/wiener_filter_via_hamiltonian.py

@@ -117,13 +117,13 @@ if __name__ == "__main__":
     # Solving the problem analytically
     m0 = D0.inverse_times(j)
 
-    sample_variance = Field(sh.domain, val=0. + 0j)
-    sample_mean = Field(sh.domain, val=0. + 0j)
+    sample_variance = Field(sh.domain, val=0.)
+    sample_mean = Field(sh.domain, val=0.)
 
     # sampling the uncertainty map
-    n_samples = 1
+    n_samples = 10
     for i in range(n_samples):
-        sample = sugar.generate_posterior_sample(m0, D0)
+        sample = fft(sugar.generate_posterior_sample(0., D0))
         sample_variance += sample**2
         sample_mean += sample
-    variance = sample_variance/n_samples - (sample_mean/n_samples)
+    variance = (sample_variance - sample_mean**2)/n_samples

nifty/basic_arithmetics.py

@@ -24,7 +24,7 @@ from .field import Field
 
 __all__ = ['cos', 'sin', 'cosh', 'sinh', 'tan', 'tanh', 'arccos', 'arcsin',
            'arccosh', 'arcsinh', 'arctan', 'arctanh', 'sqrt', 'exp', 'log',
-           'conjugate', 'clipped_exp', 'limited_exp']
+           'conjugate', 'clipped_exp', 'limited_exp', 'limited_exp_deriv']
 
 
 def _math_helper(x, function):
@@ -101,15 +101,28 @@ def clipped_exp(x):
 
 
 def limited_exp(x):
-    thr = 200
-    expthr = np.exp(thr)
-    return _math_helper(x, lambda z: _limited_exp_helper(z, thr, expthr))
-
+    return _math_helper(x, _limited_exp_helper)
+
+def _limited_exp_helper(x):
+    thr = 200.
+    mask = x>thr
+    if np.count_nonzero(mask) == 0:
+        return np.exp(x)
+    result = ((1.-thr) + x)*np.exp(thr)
+    result[~mask] = np.exp(x[~mask])
+    return result
 
-def _limited_exp_helper(x, thr, expthr):
-    mask = (x > thr)
-    result = np.exp(x)
-    result[mask] = ((1-thr) + x[mask])*expthr
+def limited_exp_deriv(x):
+    return _math_helper(x, _limited_exp_deriv_helper)
+
+def _limited_exp_deriv_helper(x):
+    thr = 200.
+    mask = x>thr
+    if np.count_nonzero(mask) == 0:
+        return np.exp(x)
+    result = np.empty_like(x)
+    result[mask] = np.exp(thr)
+    result[~mask] = np.exp(x[~mask])
     return result
 
 

nifty/field.py

@@ -330,7 +330,7 @@ class Field(Loggable, Versionable, object):
         Returns
         -------
         out : Field
-            The output object. It's domain is a PowerSpace and it contains
+            The output object. Its domain is a PowerSpace and it contains
             the power spectrum of 'self's field.
 
         See Also

nifty/minimization/line_searching/line_search.py

@@ -28,7 +28,7 @@ class LineSearch(with_metaclass(abc.ABCMeta, type('NewBase', (Loggable, object),
     """Class for determining the optimal step size along some descent direction.
 
     Initialize the line search procedure which can be used by a specific line
-    search method. Its finds the step size in a specific direction in the
+    search method. It finds the step size in a specific direction in the
     minimization process.
 
     Attributes

nifty/sugar.py

@@ -108,7 +108,7 @@ def generate_posterior_sample(mean, covariance):
     R = covariance.R
     N = covariance.N
 
-    power = S.diagonal().power_analyze()**.5
+    power = sqrt(S.diagonal().power_analyze())
     mock_signal = power.power_synthesize(real_signal=True)
 
     noise = N.diagonal(bare=True)