_5_comp.py

# The code is for Comparison
# Authors: Feng Yingxin(5692148), Gong Sicong(5711932), Rao Chengzhi(5841089)
import math
import json
import numpy as np
import rasterio
import matplotlib.pyplot as plt

num_tick = 5


def show_scatter(path, X, Y, Z):
    plt.figure(figsize=(8, 6))
    plt.scatter(x=X, y=Y, c=Z, s=0.1)
    plt.title(path.strip(".txt"))
    plt.colorbar()
    plt.clim(-5, 40)
    plt.axis("equal")
    plt.savefig('figure/' + path + '.png')
    # plt.show()


def show_image(path, result, x_positions, x_labels, y_positions, y_labels):
    plt.figure(figsize=(8, 6))
    plt.imshow(result)
    plt.colorbar()
    plt.title(path.strip('.npy'))
    plt.clim(-5, 40)
    plt.xticks(x_positions, x_labels)
    plt.yticks(y_positions, y_labels)
    plt.savefig('figure/' + path + '.png')
    # plt.show()


def show_diff(path, diff, x_positions, x_labels, y_positions, y_labels):
    plt.figure(figsize=(8, 6))
    plt.imshow(diff, cmap='Spectral_r')
    plt.colorbar()
    plt.title(path.strip('.npy'))
    plt.clim(-1, 1)
    plt.xticks(x_positions, x_labels)
    plt.yticks(y_positions, y_labels)
    plt.savefig('figure/' + path + '.png')
    # plt.show()


def get_diff(diff):
    nonan_num = np.count_nonzero(~np.isnan(diff))
    diff_squre = diff * diff
    diff_squre[np.isnan(diff_squre)] = 0
    mean = np.sum(np.sum(diff_squre)) / nonan_num
    root = math.sqrt(mean)
    return root


def comp():
    # Open the parameter file
    with open('file/parameter.json', 'r') as pfile:
        parameter = json.load(pfile)
        x_min, x_max = parameter['crop']['x_min'], parameter['crop']['x_max']
        y_min, y_max = parameter['crop']['y_min'], parameter['crop']['y_max']
        grid_size = parameter['thin']['grid_size']
        thin_rate = parameter['thin']['thin_rate']

    x_labels = np.linspace(x_min, x_max, num_tick)
    y_labels = np.linspace(y_min, y_max, num_tick)

    with rasterio.open('file\image.tif') as src:
        nodata = src.nodata
        pdok = src.read(1)
        pdok[pdok == nodata] = np.nan
        print(pdok.shape)

    np.save("file/pdok", pdok)
    pdok = np.load("file/pdok.npy")

    # Get the input parameters
    lst_input = []
    for size in grid_size:
        for rate in thin_rate:
            lst_input.append({'size': size, 'rate': rate})
    print('cloth-ground!')
    for path in lst_input:
        print(path)
        path_cloth = 'cloth_size_{}_rate_{}'.format(path['size'], path['rate'])
        path_ground = 'ground_size_{}_rate_{}'.format(path['size'], path['rate'])
        cloth = np.load('file/' + path_cloth + '.npy')
        ground = np.load('file/' + path_ground + '.npy')
        diff = cloth - ground
        x_positions = np.linspace(0, diff.shape[0], num_tick)
        y_positions = np.linspace(0, diff.shape[1], num_tick)
        show_diff('cloth-ground_size_{}_rate_{}'.format(path['size'], path['rate']), diff, x_positions, x_labels,
                  y_positions, y_labels)
        print(get_diff(diff))

    lst_path_cloth = ['cloth_size_{}_rate_{}'.format(input['size'], input['rate']) for input in lst_input]
    lst_path_ground = ['ground_size_{}_rate_{}'.format(input['size'], input['rate']) for input in
                       lst_input]
    lst_path = lst_path_cloth + lst_path_ground

    print('pdok!')
    for path in lst_path:
        print(path)
        ours = np.load('file/' + path + '.npy')
        x_positions = np.linspace(0, ours.shape[0], num_tick)
        y_positions = np.linspace(0, ours.shape[1], num_tick)
        diff = ours - pdok
        show_image('pdok_' + path, pdok, x_positions, x_labels, y_positions, y_labels)
        show_image('ours_' + path, ours, x_positions, x_labels, y_positions, y_labels)
        show_diff('diff_' + path, diff, x_positions, x_labels, y_positions, y_labels)
        print(get_diff(diff))

    print('thin!')
    for path in lst_path:
        with open("file/" + path + '.txt', "r") as ifile:
            X, Y, Z = [], [], []
            for line in ifile:
                strip = line.strip('\n')
                split = strip.split(',')
                x, y, z = float(split[0]), float(split[1]), float(split[2])
                X.append(x), Y.append(y), Z.append(z)
            show_scatter(path, X, Y, Z)


def main():
    comp()


if __name__ == "__main__":
    main()

'''
with rasterio.open('file\M_58DZ1.tif') as src:
    xmin, ymin, xmax, ymax = (190750, 351650, 191250, 352150)
    window = src.window(xmin, ymin, xmax, ymax)
    data = src.read(window=window)
    meta = src.meta.copy()
    meta.update({'height': data.shape[1], 'width': data.shape[2]})

with rasterio.open('file\image.tif', 'w', **meta) as dst:
    dst.write(data)
'''