def grid(images):
amount = len(images)
cols = math.ceil(math.sqrt(amount))
rows = math.ceil(amount / cols)
# get imgs as array of Image.Image objects
imgs = utils.images(images)
strips = []
for i in range(rows):
lower = cols * i
upper = lower + cols
if upper >= amount:
upper = amount
sham = Image.new('RGB', imgs[upper - 1].size)
dummies = [sham] * (cols * rows - amount)
strip_imgs = imgs[lower:upper] + dummies
else:
strip_imgs = images[lower:upper]
strips.append(horizontal(strip_imgs))
grid_img = vertical(strips)
return grid_img
def vertical(images):
# get imgs as array of Image.Image objects
imgs = utils.images(images)
# width of the strip is minimal width of all images
width = min([i.width for i in imgs])
resized = [resize.resize(img, width, None) for img in imgs]
# height of the strip is sum of height of all images
height = sum([img.height for img in resized])
# create new image for strip
strip = Image.new('RGB', (width, height))
# stack images by offseting them by height of previous
y_offset = 0
for i in resized:
strip.paste(i, (0, y_offset))
y_offset += i.height
return strip
def horizontal(images):
# get imgs as array of Image.Image objects
imgs = utils.images(images)
# height of the strip is minimal height of all images
height = min([i.height for i in imgs])
resized = [resize.resize(img, None, height) for img in imgs]
# width of the strip is sum of width of all images
width = sum([img.width for img in resized])
# create new image for strip
strip = Image.new('RGB', (width, height))
# stack images by offseting them by width of previous
x_offset = 0
for i in resized:
strip.paste(i, (x_offset, 0))
x_offset += i.width
return strip
title='Test Set Visualization',
xlabel='Latent dim 1',
ylabel='Latent dim 2')
print('\nCorresponding mapping coordinates in the latent space - one image per digit:\n')
encoded_x_test = encoder.predict(x_test)
digits = {}
i = 0
while len(digits) < 10:
digits[y_test[i]] = encoded_x_test[i]
i += 1
ut.output('g.c', pd.DataFrame(digits))
# (d)
z_sample = np.array([[-2.5, 0.55]])
decoded_x = generator.predict(z_sample)
ut.image(decoded_x)
ut.plt_save('g.d')
# (e)
source, target = digits[3], digits[5]
(x1, y1), (x2, y2) = source, target
a = (y2 - y1) / (x2 - x1)
b = y1 - (a * x1)
f = lambda _x_: a * _x_ + b
x_samples = np.linspace(x1, x2, num=10)
samples = [np.array([[_x, f(_x)]]) for _x in x_samples]
ut.images(map(generator.predict, samples), map(str, x_samples), n_cols=3)
ut.plt_save('g.e')