notebook.community

Edit and run



In [1]:

    
%load_ext autoreload
%autoreload 2



In [33]:

    
import os
import sys
import numpy as np
import matplotlib.pyplot as plt
import hashlib
import time
import shutil



In [3]:

    
MNM_nb_folder = os.path.join('..', '..', '..', 'side_project', 'network_builder')
sys.path.append(MNM_nb_folder)



In [4]:

    
from MNM_nb import *
import MNMAPI



In [5]:

    
data_folder = os.path.join('..', '..', '..', 'data', 'input_files_2link_fix')
new_folder = os.path.join('MNM_cache', 'input_files_2link_fix')



In [6]:

    
nb = MNM_network_builder()









    



MNM_config
MNM_pathtable



In [7]:

    
nb.load_from_folder(data_folder)



In [8]:

    
nb.config









    Out[8]:





[DTA]
network_name = Snap_graph
unit_time = 5
total_interval = 200
assign_frq = 10
start_assign_interval = 0
max_interval = 10
flow_scalar = 10
num_of_link = 4
num_of_node = 5
num_of_O = 2
num_of_D = 1
OD_pair = 2
routing_type = Fixed

[STAT]
rec_mode = LRn
rec_mode_para = 12
rec_folder = record
rec_volume = 0
volume_load_automatic_rec = 0
volume_record_automatic_rec = 0
rec_tt = 0
tt_load_automatic_rec = 0
tt_record_automatic_rec = 0

[FIXED]
path_file_name = path_table
num_path = 2
choice_portion = Buffer
route_frq = 10

Config



In [9]:

    
observed_links = [3]
estimated_paths = np.arange(0, 2)

dar



In [10]:

    
from scipy.sparse import coo_matrix
def massage_raw_dar(raw_dar, f, num_link, ass_freq, observed_links, estimated_paths, num_interval):
    num_e_path = len(estimated_paths)
    num_e_link = len(observed_links)
    link_seq = (np.array(map(lambda x: observed_links.index(x), raw_dar[:, 2].astype(np.int)))
                + raw_dar[:, 3] * num_e_link / ass_freq).astype((np.int))
    path_seq = (raw_dar[:, 0] + raw_dar[:, 1] * num_e_path).astype(np.int)
    p = raw_dar[:, 4] / f[path_seq]

    mat = coo_matrix((p, (link_seq, path_seq)), 
                   shape=(num_interval * num_e_link, num_interval * num_e_path)).toarray()
    return mat



In [11]:

    
def get_x(f, nb, observed_link_list, estimated_paths, new_folder):
    num_interval = nb.config.config_dict['DTA']['max_interval']
    ass_freq = nb.config.config_dict['DTA']['assign_frq']
    num_link = nb.config.config_dict['DTA']['num_of_link']
    total_interval = num_interval * ass_freq
    nb.config.config_dict['DTA']['total_interval'] = total_interval * 2
    num_path = nb.config.config_dict['FIXED']['num_path']
    nb.update_demand_path(f)
    nb.dump_to_folder(new_folder)
    a = MNMAPI.dta_api()
    a.initialize(new_folder)
    a.register_links(observed_link_list)
    a.register_paths(estimated_paths)
    a.install_cc()
    a.install_cc_tree()
    a.run_whole()
    x = a.get_link_inflow(np.arange(0, total_interval, ass_freq), np.arange(0, total_interval, ass_freq) + ass_freq).reshape((len(observed_link_list), total_interval/ass_freq), order = 'F').flatten()
    raw_dar = a.get_dar_matrix(np.arange(0, total_interval, ass_freq), np.arange(0, total_interval, ass_freq) + ass_freq)
    dar = massage_raw_dar(raw_dar, f, num_link, ass_freq, observed_link_list, estimated_paths, num_interval)
    return x, dar

actual OD



In [12]:

    
total_interval = 100
ass_freq = 10
num_path = nb.config.config_dict['FIXED']['num_path']
num_link = nb.config.config_dict['DTA']['num_of_link']
num_interval = nb.config.config_dict['DTA']['max_interval']



In [13]:

    
f_real = np.random.rand(num_path * num_interval) * 10
x_real, dar = get_x(f_real, nb, observed_links, estimated_paths, new_folder)



In [14]:

    
x_real









    Out[14]:





array([ 10.2,   9.4,  12.3,  10.6,   6. ,   8.4,   9.5,   5.6,  18.7,  15.1])



In [15]:

    
# nb.update_demand_path(f_real)



In [16]:

    
# nb.dump_to_folder(new_folder)



In [29]:



In [35]:

    
hash = hashlib.sha1()
hash.update(str(time.time()))
new_folder = str(hash.hexdigest())
nb.update_demand_path(f_real)
nb.dump_to_folder(new_folder)
a = MNMAPI.dta_api()
a.initialize(new_folder)
shutil.rmtree(new_folder)
a.register_links([3])
a.register_paths(np.arange(0, 2))
a.install_cc()
a.install_cc_tree()
a.run_whole()









    Out[35]:





0L



In [32]:



In [27]:









    



3a8dc14d8397ac668c13e28c2303949e494ee015
3a8dc14d83



In [18]:

    
a.get_dar_matrix(np.arange(0, total_interval, ass_freq), np.arange(0, total_interval, ass_freq) + ass_freq)









    Out[18]:





array([[  1. ,   0. ,   3. ,   0. ,   2. ],
       [  0. ,   0. ,   3. ,   0. ,   8.2],
       [  1. ,   1. ,   3. ,  10. ,   7.7],
       [  0. ,   1. ,   3. ,  10. ,   1.7],
       [  1. ,   2. ,   3. ,  20. ,   7.9],
       [  0. ,   2. ,   3. ,  20. ,   4.4],
       [  1. ,   3. ,   3. ,  30. ,   7.5],
       [  0. ,   3. ,   3. ,  30. ,   3.1],
       [  1. ,   4. ,   3. ,  40. ,   1.5],
       [  0. ,   4. ,   3. ,  40. ,   4.3],
       [  1. ,   5. ,   3. ,  50. ,   2.2],
       [  0. ,   5. ,   3. ,  50. ,   6.2],
       [  1. ,   6. ,   3. ,  60. ,   5.2],
       [  0. ,   6. ,   3. ,  60. ,   4.3],
       [  1. ,   7. ,   3. ,  70. ,   5.1],
       [  0. ,   7. ,   3. ,  70. ,   0.5],
       [  1. ,   8. ,   3. ,  80. ,   9.3],
       [  0. ,   8. ,   3. ,  80. ,   9.4],
       [  1. ,   9. ,   3. ,  90. ,   8.7],
       [  0. ,   9. ,   3. ,  90. ,   6.5]])



In [19]:

    
# a.get_cur_loading_interval()

Estimate



In [20]:

    
f_e = np.random.rand(num_path * num_interval)*5
x_e, dar_e = get_x(f_e, nb, observed_links, estimated_paths, new_folder)



In [21]:

    
step_size = 0.1



In [22]:

    
for i in range(50):
    x_e, dar = get_x(f_e, nb, observed_links, estimated_paths, new_folder)
    grad = - dar.T.dot(x_real - x_e)
    print np.linalg.norm(x_e - x_real), np.linalg.norm(f_e - f_real)
    f_e -= grad * step_size / np.sqrt(i + 1)









    



23.5376294473 19.8221021911
18.8098378515 17.1012877641
16.1576607218 15.7290242259
14.302447343 14.8251908409
12.8965111561 14.1780141447
11.8156675647 13.6889887948
10.8013887996 13.3078558297
9.90656348084 13.0030588674
9.22117129219 12.7573857219
8.6197447758 12.5538310467
8.13265024454 12.3845649729
7.56967634711 12.2405250131
7.18401002226 12.1197252491
6.81322243876 12.0154093914
6.46219776856 11.9258241172
6.10573500899 11.8481695295
5.92030404625 11.7809503382
5.54256258422 11.7209728028
5.26497863244 11.5408047449
5.08527285404 11.4883752648
4.77912125814 11.4399022098
4.6357307946 11.3973191703
4.38178046004 11.3589897755
4.20119030752 11.3248535897
4.05215991787 11.2941844646
3.92428337407 11.2665277787
3.68374809128 11.2414682735
3.63455636908 11.2190599532
3.45832329316 11.1976736378
3.27261363439 11.1785190687
3.26190128606 11.1612369897
3.17490157328 11.1450025941
2.99666481275 11.1301888161
2.80356915378 11.1167586735
2.64575131106 11.1048970135
2.683281573 11.0941871921
2.65894716006 11.0835685855
2.54558441227 11.07388492
2.4289915603 11.0646384722
2.34946802489 11.0562914922
2.29128784748 11.0488364316
2.16794833887 11.0419466893
2.2 11.0356456807
1.97230829233 11.0292470034
2.04939015319 11.0238410877
1.89472953215 11.0183876323
1.93132079158 11.0135237313
1.788854382 11.0088067916
1.788854382 11.0046039872
1.75783958312 11.0005590351



In [23]:

    
f_e









    Out[23]:





array([ 4.27074717,  5.77188792,  5.27645651,  3.71264918,  6.82524988,
        4.81834652,  5.01495382,  5.45732112,  0.59254426,  5.18963998,
        5.06661818,  2.99915968,  4.4734873 ,  4.6135044 ,  2.20526426,
        3.58759296,  7.93295609,  9.64893005,  5.15826192,  9.21908952])



In [24]:

    
f_real









    Out[24]:





array([ 8.15804598,  1.95560978,  1.65903667,  7.65930837,  4.34551743,
        7.94650391,  3.09611815,  7.46724298,  4.31241003,  1.52434044,
        6.17122785,  2.2181613 ,  4.25540151,  5.14555835,  0.44283175,
        5.11350549,  9.37581069,  9.36942636,  6.44685073,  8.72100766])



In [26]:

    
plt.plot(f_e, f_real, '*')
plt.show()