data_visualisation.py

import numpy as np
import matplotlib.pyplot as plt
from util import *

datasets = [np.load('training_data/{}.npy'.format(pth))
        for pth in ['Phantom', 'FFBP128', 'Sinogram']]



# plot some images
plot = False

if plot:
    nimgs = 2
    fig, ax = plt.subplots(2, nimgs)
    for i, data in enumerate(datasets):
        for j in range(nimgs):
            ax[i, j].imshow(data[j, 0], vmin=0.15)
    plt.tight_layout()

# baseline MSE
EC = False
if EC:
    phant = datasets[0]
    img = one_hot_y(torch.tensor(phant[:10]))
    fbp = prep_img(datasets[1])
    #
    plt.figure()
    diff = (((fbp - phant)**2)).mean(axis=(1, 2, 3))**0.5
    plt.hist(diff)
import scipy.ndimage as nd
# sino = datasets[-1][0,0]0
# img = np.zeros([1024, 1024])
# for a, l in enumerate(sino):
#     print(a)
#     lay = np.tile(l, (1024, 1))
#     lay = nd.rotate(lay, a*0.7)
#     mm = lay.shape[0] - 1024
#     if mm!=0:
#         edge = mm//2
#         lay = lay[edge:-(mm-edge), edge:-(mm-edge)]
#     img+=lay