Added Utilities ZeroPad

proxtoolbox/Utilities/ZeroPad.py

+#######################################################################
+#                                                                    
+#                             ZeroPad.m
+#                            written by
+#                           Russell Luke
+#                      Nachwuchsforschergruppe
+#                   Inst. fuer Num. u. Angew. Math
+#                       Universitaet Goettingen
+#                          August 12, 2001
+# DESCRIPTION: Function for zero padding an array preparatory to
+# using the FFT.  Automatically resizes the array to the next
+# largest diad  by adding np.zeros symmetrically to the array and again
+# symmetrically doubles the size of the array.  Takes one or two
+# dimensional arrays.
+#
+# INPUT:
+#              m = data array
+#
+# OUTPUT:      padded_m = the zero padded array ready for use with
+#                         the FFT.
+#
+# USAGE:       padded_m = ZeroPad(m)
+# Non-standard function calls: 
+#
+########################################################################
+import numpy as np
+
+def ZeroPad(m):
+
+    dim = np.shape(m)
+    print(dim)
+    orient = np.argmax(dim)
+    major = dim[orient]
+    minor = np.min(dim)
+    tmp = np.log2(major+1)
+    n = np.ceil(tmp)
+    tmp = np.round(2**n-(2**tmp-1))
+    if np.mod(tmp,2)==1:
+      leftpad =  int((tmp+1)/2)
+      rightpad = int((tmp-1)/2)
+    else:
+      leftpad =  int(tmp/2)
+      rightpad = leftpad
+
+    if orient==2:
+      tmp = np.zeros((minor,leftpad))
+      padded_m = np.concatenate((tmp,m),axis=1)
+      tmp = np.zeros((minor,rightpad))
+      padded_m = np.concatenate((padded_m,tmp),axis=1)
+    else:
+      tmp = np.zeros((leftpad,minor))
+      print(tmp)
+      padded_m = np.concatenate((tmp,m),axis=0)
+      tmp = np.zeros((rightpad,minor))
+      padded_m = np.concatenate((padded_m,tmp),axis=0)
+
+    major = 2**n
+
+    if minor!=1:
+      tmp = np.log2(minor)
+      tmp = np.round(2**n-2**tmp)
+      if np.mod(tmp,2)==1:
+        leftpad =  (tmp+1)/2
+        rightpad = (tmp-1)/2
+      else:
+        leftpad =  tmp/2
+        rightpad = leftpad
+      if orient==2:
+        tmp = np.zeros((leftpad,major))
+        padded_m = np.concatenate((tmp,padded_m),axis=0)
+        tmp = np.zeros((rightpad,major))
+        padded_m = np.concatenate((padded_m,tmp),axis=0)
+      else:
+        tmp = np.zeros((major,leftpad))
+        padded_m = np.concatenate((tmp,padded_m),axis=1)
+        tmp = np.zeros((major,rightpad))
+        padded_m = np.concatenate((padded_m,tmp),axis=1)
+
+    return padded_m
+