| # Copyright 2016 The Vanadium Authors. All rights reserved. |
| # Use of this source code is governed by a BSD-style |
| # license that can be found in the LICENSE file. |
| |
| """Creates HTML files to visualize the images in different clusters.""" |
| |
| import argparse |
| import json |
| import numpy as np |
| import os |
| |
| from random import randint |
| |
| |
| def write_html(img_names, k, label, data_folder_path): |
| """Creates a HTML file for each cluster.""" |
| |
| folder_name = os.path.join('cluster_viz', str(k)) |
| if not os.path.exists(folder_name): |
| os.makedirs(folder_name) |
| |
| html_filename = os.path.join(folder_name, str(label) + '.html') |
| # If the file exists, delete it. |
| if os.path.exists(html_filename): |
| os.remove(html_filename) |
| |
| # Many screens form the a session would be the same. We want to show only |
| # one of those screens. So we use the list of unique views from each session. |
| with open('unique_views.json') as namefile: |
| unique_views = json.loads(namefile.read()) |
| |
| with open(html_filename, 'a') as html_file: |
| html_file.write('<!DOCTYPE html>\n') |
| html_file.write('<html>\n') |
| html_file.write(' <head>\n') |
| html_file.write(' <title>Clusters</title>\n') |
| html_file.write(' </head>\n') |
| html_file.write(' <body>\n') |
| |
| html_file.write(' <div>') |
| # We provide links at the top of the HTML file to go to the previous, next, |
| # or a random cluster. This helps when visually inspecting clusters. |
| if label != 1: |
| html_file.write((' <a style="padding-right:30px" href="./{}.html">' |
| 'Previous</a>').format(str(label - 1))) |
| html_file.write((' <a style="padding-right:30px" href="./{}.html">' |
| 'Random</a>').format(str(randint(1, k)))) |
| if label != k: |
| html_file.write(' <a align="right" href="./{}.html">Next</a>'.format( |
| str(label + 1))) |
| html_file.write(' </div>') |
| |
| for img_name in img_names: |
| session_id, img = img_name.split('_') |
| views = unique_views.get(str(session_id), []) |
| if views and img.split('.')[0] in views: |
| html_file.write(' <div style="vertical-align: top; display:' |
| ' inline-block; text-align: center;">') |
| html_file.write((' <img alt={} src="{}/{}/img/{}" style="height: ' |
| '300px; padding: 0px;">\n').format( |
| str(img_name), data_folder_path, session_id, img)) |
| html_file.write(' </div>') |
| html_file.write(' </body>\n') |
| html_file.write('</html>\n') |
| |
| if __name__ == '__main__': |
| parser = argparse.ArgumentParser() |
| parser.add_argument('data_folder_path', |
| help=('Full path to the data folder. It could contain ' |
| 'multiple folders for different sessions inside it.' |
| ) |
| ) |
| args = parser.parse_args() |
| data_folder_path = args.data_folder_path |
| |
| ks = [2**exp for exp in range(5, 11)] |
| with open('image_names.json') as namefile: |
| image_names = json.loads(namefile.read())['img_names'] |
| |
| print len(image_names) |
| |
| for k in ks: |
| print 'K: ' + str(k) |
| labels = np.load('clusters/labels_' + str(k) + '.npy') |
| for label in range(1, k + 1): |
| indices = [i for i, j in enumerate(labels) if j == label] |
| img_names = [image_names[idx] for idx in indices] |
| write_html(img_names, k, label, data_folder_path) |
| |
