Spaces:
Running
Running
| import sys | |
| sys.path.append('../XGBoost_Prediction_Model/') | |
| import warnings | |
| warnings.filterwarnings("ignore") | |
| import Predict | |
| import torch | |
| import numpy as np | |
| import os | |
| from os.path import isfile, isdir, join | |
| from Magazine_Optimization import * | |
| import cv2 as cv | |
| import time | |
| mypath = '../XGBoost_Prediction_Model/Magazine_Optimization_Demo/Magazines' | |
| # results = torch.load('../XGBoost_Prediction_Model/Magazine_Optimization_Demo/results') | |
| results = torch.load('../XGBoost_Prediction_Model/Magazine_Optimization_Demo/results_CNN') | |
| for f in os.listdir(mypath): | |
| if isdir(join(mypath, f)): | |
| print('Currently processing Magazine '+f+'......') | |
| path_temp = join(mypath, f) | |
| dir_list = [] | |
| for sub_f in os.listdir(path_temp): | |
| if isdir(join(path_temp, sub_f)): | |
| sub_path_temp = join(path_temp, sub_f) | |
| if (sub_f.split('_')[0]) == 'Jpg': | |
| dir_list = os.listdir(sub_path_temp) | |
| dir_list.sort() | |
| for i in range(len(dir_list)): | |
| dir_list[i] = join(sub_path_temp,dir_list[i]) | |
| else: | |
| Slots = torch.load(join(sub_path_temp,'Slots')).astype('int32') | |
| strategies = results[f]['AG'][0].split('; ')[:-1] | |
| newpath = '../XGBoost_Prediction_Model/Magazine_Optimization_Demo/Optimal_Magazines_CNN/'+f | |
| if not os.path.exists(newpath): | |
| os.makedirs(newpath) | |
| for item in strategies: | |
| ad_id = int(item.split(' ')[2]) | |
| ctpg_id = int(item.split(' ')[5]) | |
| ad_loc = 1-Slots[ad_id] | |
| ctpg_loc = Slots[ctpg_id] | |
| ad = cv.imread(dir_list[ad_id])[:,(ad_loc*640):((ad_loc+1)*640),:] | |
| ctpg = cv.imread(dir_list[ctpg_id])[:,(ctpg_loc*640):((ctpg_loc+1)*640),:] | |
| img = np.concatenate((ad,ctpg),axis=1) | |
| cv.imwrite(newpath+'/Newly_Assigned_Page'+str(ad_id)+'.jpg', img) | |