OptiMask: Figures Generation

# ! pip install optimask pyoxipng

from pathlib import Path

import matplotlib.pyplot as plt
import numpy as np
from optimask import OptiMask
from optimask.utils import generate_mar, plot
from oxipng import optimize

[p.unlink() for p in Path('.').glob('*.png')]

DPI = 200

x = np.zeros((12, 5))
x[7, [1, 3]] = np.nan

plt.figure()
plt.subplot(1, 3, 1)
plot(x, show=False, title="original data")
plt.subplot(1, 3, 2)
plot(x, rows_to_remove=[7], show=False, title="remove row")
plt.subplot(1, 3, 3)
plot(x, cols_to_remove=[1, 3], show=False, title="remove columns")
plt.savefig('at_hand_two.png', bbox_inches='tight', dpi=DPI)
plt.show()

def heights(x, axis=0):
    return (x.shape[axis] - np.argmax(np.flip(x, axis=axis), axis=axis)).astype(np.uint32)


def to_plot(x):
    xp = x.astype(float)
    xp[xp > 0] = np.nan
    return xp

m, n = 40, 30
x = generate_mar(m=m, n=n, ratio=0.04, rng=1)
x[12:19, 10] = np.nan

plot(x, show=False)
plt.savefig('algo_data.png', bbox_inches='tight', dpi=DPI)
plt.show()


nan_rows, nan_cols = np.isnan(x).nonzero()
nan_rows, nan_cols = np.unique(nan_rows), np.unique(nan_cols)
xp = np.isnan(x[nan_rows][:, nan_cols])
hx, hy = heights(xp, axis=0), heights(xp, axis=1)
plot(to_plot(xp), xticks=hx, yticks=hy, show=False)
plt.savefig('algo_0.png', bbox_inches='tight', dpi=DPI)
plt.show()

hx, hy = heights(xp, axis=0).astype(np.uint32), heights(xp, axis=1).astype(np.uint32)
k = 0
plt.figure(figsize=(12, 5))
while not (OptiMask.is_decreasing(hx) and OptiMask.is_decreasing(hy)):
    axis = k % 2
    p_step = np.argsort(-heights(xp, axis=axis))
    if axis == 0:
        xp = xp[:, p_step]
    if axis == 1:
        xp = xp[p_step]
    hx, hy = heights(xp, axis=0).astype(np.uint32), heights(xp, axis=1).astype(
        np.uint32
    )
    plt.subplot(1, 3, k+1)
    plot(to_plot(xp), xticks=hx, yticks=hy, title=f"Iteration {k+1}", show=False)
    k += 1
plt.tight_layout()
plt.savefig('algo_iterations.png', bbox_inches='tight', dpi=DPI)
plt.show()

def rectangle(xmin, ymin, xmax, ymax, c='k', lw=0.4):
    x = [xmin, xmax, xmax, xmin, xmin]
    y = [ymin, ymin, ymax, ymax, ymin]
    plt.plot(x, y, c=c, linestyle='--', lw=lw)

opt_area = 0
plot(np.zeros_like(x), show=False)
plot(to_plot(xp), xticks=hx, yticks=hy, show=False)
for i0 in range(len(hx)):
    if (i0 == 0) or (i0 > 0 and (hx[i0] < hx[i0-1])):
        rectangle(i0, hx[i0], n, m)
        if (n-i0)*(m-hx[i0]) > opt_area:
            opt_area, opt_index = (n-i0)*(m-hx[i0]), i0
rectangle(len(hx), 0, n, m)
rectangle(opt_index, hx[opt_index], n, m, c='r', lw=1)
plt.savefig('algo_result_permuted_space.png', bbox_inches='tight', dpi=DPI)
plt.show()

rows, cols = OptiMask(n_tries=10000).solve(x)

plot(x, rows, cols, show=False)
plt.savefig('algo_result.png', bbox_inches='tight', dpi=DPI)
plt.show()

[optimize(path) for path in Path('.').glob('*.png')]