from numpy import ones, round, zeros, expand_dims, Inf, tile, arange, repeat, array
from functools import wraps
from matplotlib import pyplot as plt
import matplotlib.cm as cm
from numpy.ma import masked_where
from numpy import maximum, minimum
import cvxpy as cp
def pplot(As, titles):
# setup
try: vmin = min([A.min() for A, t in zip(As[:-1], titles) if "missing" not in t]) # for pixel color reference
except: vmin = As[0].min()
try: vmax = max([A.max() for A, t in zip(As[:-1], titles) if "missing" not in t])
except: vmax = As[0].max()
my_dpi = 96
plt.figure(figsize=(1.4*(250*len(As))/my_dpi, 250/my_dpi), dpi = my_dpi)
for i, (A, title) in enumerate(zip(As, titles)):
plt.subplot(1, len(As), i+1)
if i == len(As)-1: vmin, vmax = A.min(), A.max()
if "missing" in title:
missing = A
masked_data = ones(As[i-1].shape)
for j,k in missing: masked_data[j,k] = 0
masked_data = masked_where(masked_data > 0.5, masked_data)
plt.imshow(As[i-1], interpolation = 'nearest', vmin = vmin, vmax = vmax)
plt.colorbar()
plt.imshow(masked_data, cmap = cm.binary, interpolation = "nearest")
else:
plt.imshow(A, interpolation = 'nearest', vmin = vmin, vmax = vmax)
plt.colorbar()
plt.title(title)
plt.axis("off")
plt.show()
#
# def unroll_missing(missing, ns):
# missing_unrolled = []
# for i, (MM, n) in enumerate(zip(missing, ns)):
# for m in MM:
# n2 = m[1] + sum([ns[j] for j in range(i)])
# missing_unrolled.append((m[0], n2))
# return missing_unrolled
#
def shrinkage(a, kappa):
""" soft threshold with parameter kappa). """
try: return maximum(a - kappa(ones(a.shape), 0)) - maximum(-a - kappa*ones(a.shape), 0)
except: return max(a - kappa, 0) - max(-1 - kappa, 0)
