Temporarily roll-back to old get_source function.

Author: sfwang

lib/utils/utils.py

@@ -131,45 +131,54 @@ def get_mesh(x_mesh, y_mesh):
     mesh = np.array([Xsim,Ysim,np.zeros(d)]);
     return mesh
 
-def get_source(n_sources=[40, 60],
-               p_n_sources=[.5, .5],
-               depth=[.1],
-               p_depth=[1],
-               amplitude=[-1,1],
-               p_amplitude=[.5,.5]):
-    """
-    Parameters:
-    -----------
-    n_sources: numpy array (int)
-    the number of sources in the plane
-    
-    p_n_sources: numpy array
-    The probabily distribution over the number of sources.
-    
-    depth: numpy array
-    the depth of sources
-    
-    p_depth: numpy array
-    The probabily distribution over the depth of sources.
-    
-    amplitude: numpy array
-    amplitude of sources.
-    
-    p_amplitude: numpy array
-    The probabily distribution over the amplitude of sources.
-    
-    Returns:
-    -------
-    sources: numpy array of shape (4, n)
-    the bio information of sources.
-    """
-    
-    number_of_sources = np.random.choice(a=n_sources, size=1, p=p_n_sources)[0]
-    sources = np.zeros(4, number_of_sources);
-    d = np.expand_dims(np.random.choice(a=depth, size=number_of_sources, p=p_depth),0)
-    loc = np.random.rand(2, number_of_sources)
-    amp = np.expand_dims(np.random.choice(a=amplitude, size=number_of_sources, p=p_amplitude),0)
-    sources = np.append(np.append(amp, loc, 0), d,0)
+# def get_source(n_sources=[40, 60],
+#                p_n_sources=[.5, .5],
+#                depth=[.1],
+#                p_depth=[1],
+#                amplitude=[-1,1],
+#                p_amplitude=[.5,.5]):
+#     """
+#     Parameters:
+#     -----------
+#     n_sources: numpy array (int)
+#     the number of sources in the plane
+#
+#     p_n_sources: numpy array
+#     The probabily distribution over the number of sources.
+#
+#     depth: numpy array
+#     the depth of sources
+#
+#     p_depth: numpy array
+#     The probabily distribution over the depth of sources.
+#
+#     amplitude: numpy array
+#     amplitude of sources.
+#
+#     p_amplitude: numpy array
+#     The probabily distribution over the amplitude of sources.
+#
+#     Returns:
+#     -------
+#     sources: numpy array of shape (4, n)
+#     the bio information of sources.
+#     """
+#
+#     number_of_sources = np.random.choice(a=n_sources, size=1, p=p_n_sources)[0]
+#     sources = np.zeros(4, number_of_sources);
+#     d = np.expand_dims(np.random.choice(a=depth, size=number_of_sources, p=p_depth),0)
+#     loc = np.random.rand(2, number_of_sources)
+#     amp = np.expand_dims(np.random.choice(a=amplitude, size=number_of_sources, p=p_amplitude),0)
+#     sources = np.append(np.append(amp, loc, 0), d,0)
+#     return sources
+
+def get_source(depth, n_sources, var_noise):
+
+    sources = np.random.rand(4, n_sources);
+    sources[3, :] = depth
+    sources[0, :] = 2*np.floor(2*sources[0, :])-1;
+    # image = evalpotential(mesh, sources);
+    # image = image.reshape((y_mesh, x_mesh)) + var_noise*np.random.randn(y_mesh, x_mesh)
     return sources
 
 def generate_heatmaps(keypoints, im_height, im_width):