sys.path.append('../../../optimaltransport')
import numpy as np
import matplotlib.pyplot as plt
from optrans.continuous import MultiVOT2D
from optrans.datasets import adni
"""
Compute the transport map between two images using the multi-scale variational
optimal transport method.
Liam Cattell -- May 2018
"""
# Load images
X, _ = adni.load_data()
img0 = adni.load_img0()
img1 = X[1]
# Multi-scale Variational OT
n_scales = 3
vot = MultiVOT2D(n_scales=n_scales, lr=0.01, alpha=0., max_iter=300, verbose=2)
img1_hat = vot.forward(img0, img1)
# Reconstruct images using transport map
img0_recon = vot.apply_forward_map(vot.transport_map_, img1)
img1_recon = vot.apply_inverse_map(vot.transport_map_, img0)
# Plot cost function for each scale
fig1, ax1 = plt.subplots(n_scales, 3)
for i, a in enumerate(ax1.reshape((-1, 3))):
import sys
sys.path.append('../../../optimaltransport')
import numpy as np
import matplotlib.pyplot as plt
from sklearn.decomposition import PCA
from optrans.datasets import adni
from optrans.visualization import plot_mode_image
"""
Perform PCA on image data, and plot a PCA mode of variation as a single image.
Liam Cattell -- March 2018
"""
# Load some image data
X, y = adni.load_data()
# Reshape data to be n_imgs-by-n_features
n_imgs, h, w = X.shape
X = X.reshape((n_imgs, h * w))
# Fit PCA transform
pca = PCA(n_components=5)
X_pca = pca.fit_transform(X)
# Plot mode of variation image
ax = plot_mode_image([pca], component=0, shape=(h, w), n_std=3, n_steps=5)
plt.show()