#!/usr/bin/env python3 import sys import os import time import random import tempfile import argparse import socket import json import numpy import matplotlib.mlab as mlab import matplotlib.pyplot as plt VERSION = "{DEVELOPMENT}" if VERSION == "{DEVELOPMENT}": script_dir = '.' try: script_dir = os.path.dirname(os.path.realpath(__file__)) except: try: script_dir = os.path.dirname(os.path.abspath(sys.argv[0])) except: pass sys.path.append("%s/../lib" % script_dir) from nanownlib import * from nanownlib.stats import * import nanownlib.storage parser = argparse.ArgumentParser( description="") parser.add_argument('db_file', default=None, help='') options = parser.parse_args() db = nanownlib.storage.db(options.db_file) def differences(db, unusual_case, rtt_type='packet'): ret_val = [s['unusual_'+rtt_type]-s['other_'+rtt_type] for s in db.subseries('train', unusual_case)] ret_val += [s['unusual_'+rtt_type]-s['other_'+rtt_type] for s in db.subseries('test', unusual_case)] return ret_val def null_differences(db, unusual_case, rtt_type='packet'): ret_val = [s['unusual_'+rtt_type]-s['other_'+rtt_type] for s in db.subseries('train_null', unusual_case)] return ret_val def timeSeries(db, probe_type, unusual_case): cursor = db.conn.cursor() query=""" SELECT time_of_day,packet_rtt AS uc,(SELECT avg(packet_rtt) FROM probes,analysis WHERE analysis.probe_id=probes.id AND probes.test_case!=:unusual_case AND probes.type=:probe_type AND sample=u.sample) AS oc FROM (SELECT time_of_day,probes.sample,packet_rtt FROM probes,analysis WHERE analysis.probe_id=probes.id AND probes.test_case =:unusual_case AND probes.type=:probe_type) u """ params = {"probe_type":probe_type,"unusual_case":unusual_case} cursor.execute(query, params) for row in cursor: yield {'time_of_day':row['time_of_day'],unusual_case:row['uc'],'other_cases':row['oc']} #samples,derived,null_derived = parse_data(input1) #trust = trustValues(derived, sum) #weights = linearWeights(derived, trust, 0.25) #print('(test): %f' % weightedMean(derived,weights)) diffs = list(differences(db, 'long')) reported_diffs = list(differences(db, 'long', 'reported')) #shorts = [s['packet_rtt'] for s in samples.values() if s['test_case']=='short'] #longs = [s['packet_rtt'] for s in samples.values() if s['test_case']=='long'] short_overtime = [(sample['time_of_day'],sample['short']) for sample in timeSeries(db,'train','short')] long_overtime = [(sample['time_of_day'],sample['long']) for sample in timeSeries(db,'train','long')] diff_overtime = [(sample['time_of_day'],sample['long']-sample['other_cases']) for sample in timeSeries(db,'train','long')] short_overtime.sort() long_overtime.sort() diff_overtime.sort() print('packet_rtt diff mean: %f' % statistics.mean(diffs)) print('packet_rtt diff median: %f' % statistics.median(diffs)) print('packet_rtt diff midhinge: %f' % midsummary(diffs)) print('packet_rtt diff trimean: %f' % trimean(diffs)) print('packet_rtt diff quadsummary: %f' % quadsummary(diffs)) print('packet_rtt diff ubersummary: %f' % ubersummary(diffs)) print('packet_rtt diff septasummary: %f' % septasummary(diffs)) print('packet_rtt diff MAD: %f' % mad(diffs)) try: print('reported diff trimean: %f' % trimean(reported_diffs)) print('reported diff quadsummary: %f' % quadsummary(reported_diffs)) print('reported diff ubersummary: %f' % ubersummary(reported_diffs)) print('reported diff septasummary: %f' % septasummary(reported_diffs)) print('reported diff MAD: %f' % mad(reported_diffs)) #import cProfile #start = time.time() #kresults = kfilter({},diffs) #print('packet_rtt diff kfilter: ', numpy.mean(kresults['est']), kresults['var']) #print('packet_rtt diff kfilter: ', kresults['est'][-1], kresults['var'][-1]) #kresults = kfilter({},reported_diffs) #print('reported diff kfilter: ', numpy.mean(kresults['est']), kresults['var'][-1]) #print('reported diff kfilter: ', kresults['est'][-1], kresults['var'][-1]) #print("kfilter time: %f" % (time.time()-start)) except: pass #print('tsval diff mean: %f' % numpy.mean(differences(db, 'long', 'tsval'))) #print('tsval null diff mean: %f' % numpy.mean(null_differences(db, 'long', 'tsval'))) #print('tsval diff weighted mean: %f' % tsvalwmean(db.subseries('train','long')+db.subseries('test','long'))) #print('tsval null diff weighted mean: %f' % tsvalwmean(db.subseries('train_null','long'))) def testKalman4D(params=None): from pykalman import KalmanFilter train = db.subseries('train','long', offset=0) test = db.subseries('test','long', offset=0) null = db.subseries('train_null','long', offset=0) measurements = numpy.asarray([(s['unusual_packet'],s['other_packet'],s['unusual_tsval'],s['other_tsval']) for s in (train+test)]) null_measurements = numpy.asarray([(s['unusual_packet'],s['other_packet'],s['unusual_tsval'],s['other_tsval']) for s in null]) if params == None: kf = KalmanFilter(n_dim_obs=4, n_dim_state=4, initial_state_mean=[quadsummary([s['unusual_packet'] for s in train]), quadsummary([s['other_packet'] for s in train]), numpy.mean([s['unusual_tsval'] for s in train]), numpy.mean([s['other_tsval'] for s in train])]) kf = KalmanFilter(n_dim_obs=4, n_dim_state=4) start=time.time() kf = kf.em(measurements[0:len(train)]+null_measurements[0:50000], n_iter=10, em_vars=('transition_matrices', 'observation_matrices', 'transition_offsets', 'observation_offsets', 'transition_covariance', 'observation_covariance', 'initial_state_mean', 'initial_state_covariance')) params = {'transition_matrices': kf.transition_matrices.tolist(), 'observation_matrices': kf.observation_matrices.tolist(), 'transition_offsets': kf.transition_offsets.tolist(), 'observation_offsets': kf.observation_offsets.tolist(), 'transition_covariance': kf.transition_covariance.tolist(), 'observation_covariance': kf.observation_covariance.tolist(), 'initial_state_mean': kf.initial_state_mean.tolist(), 'initial_state_covariance': kf.initial_state_covariance.tolist()} print("Learned Params:\n") import pprint pprint.pprint(params) print("pykalman em time: %f" % (time.time()-start)) #kf = KalmanFilter(n_dim_obs=2, n_dim_state=2, **params) num_obs=5000 for offset in range(50000,100000+num_obs,num_obs): start=time.time() m = measurements[offset:offset+num_obs] #params['initial_state_mean']=[quadsummary([s[0] for s in m]), # quadsummary([s[1] for s in m]), # numpy.mean([s[2] for s in m]), # numpy.mean([s[3] for s in m])] kf = KalmanFilter(n_dim_obs=4, n_dim_state=4, **params) (smoothed_state_means, smoothed_state_covariances) = kf.smooth(m) #print("pykalman smooth time: %f" % (time.time()-start)) up = numpy.mean([m[0] for m in smoothed_state_means]) op = numpy.mean([m[1] for m in smoothed_state_means]) #print("packet_rtt pykalman final:", smoothed_state_means[-1][0]-smoothed_state_means[-1][1]) print("packet_rtt pykalman mean:", up-op) print("packet_rtt mean:", numpy.mean([s[0]-s[1] for s in m])) #up = numpy.mean([m[2] for m in smoothed_state_means]) #op = numpy.mean([m[3] for m in smoothed_state_means]) #print("tsval_rtt pykalman final:", smoothed_state_means[-1][2]-smoothed_state_means[-1][3]) #print("tsval_rtt pykalman mean:", up-op) #print("tsval_rtt mean:", numpy.mean([s[2]-s[3] for s in m])) for offset in range(0,len(null_measurements)+num_obs,num_obs): start=time.time() m = null_measurements[offset:offset+num_obs] #params['initial_state_mean']=[quadsummary([s[0] for s in m]), # quadsummary([s[1] for s in m]), # numpy.mean([s[2] for s in m]), # numpy.mean([s[3] for s in m])] kf = KalmanFilter(n_dim_obs=4, n_dim_state=4, **params) (smoothed_state_means, smoothed_state_covariances) = kf.smooth(m) up = numpy.mean([m[0] for m in smoothed_state_means]) op = numpy.mean([m[1] for m in smoothed_state_means]) #print("null packet_rtt pykalman final:", smoothed_state_means[-1][0]-smoothed_state_means[-1][1]) print("null packet_rtt pykalman mean:", up-op) print("null packet_rtt mean:", numpy.mean([s[0]-s[1] for s in m])) #up = numpy.mean([m[2] for m in smoothed_state_means]) #op = numpy.mean([m[3] for m in smoothed_state_means]) #print("null tsval_rtt pykalman final:", smoothed_state_means[-1][2]-smoothed_state_means[-1][3]) #print("null tsval_rtt pykalman mean:", up-op) #print("null tsval_rtt mean:", numpy.mean([s[2]-s[3] for s in m])) echo_vm_5k={'initial_state_covariance': [[33599047.5, -18251285.25, 3242535690.59375, -8560730487.84375], [-18251285.25, 9914252.3125, -1761372688.59375, 4650260880.1875], [3242535690.59375, -1761372688.59375, 312926663745.03125, -826168494791.7188], [-8560730487.84375, 4650260880.1875, -826168494791.7188, 2181195982530.4688]], 'initial_state_mean': [12939012.5625, 12934563.71875, 13134751.608, 13138990.9985], 'observation_covariance': [[11960180434.411114, 4760272534.795976, 8797551081.431936, 6908794128.927051], [4760272534.795962, 12383598172.428213, 5470747537.2599745, 11252625555.297853], [8797551081.431955, 5470747537.2601185, 1466222848395.7058, 72565713883.12643], [6908794128.927095, 11252625555.297981, 72565713883.12654, 1519760903943.507]], 'observation_matrices': [[1.4255288693095167, -0.4254638445329988, 0.0003406844036817347, -0.0005475021956726778], [-0.46467270827589857, 1.4654311778340343, -0.0003321330280128265, -0.0002853945703691352], [-0.2644570970067974, -0.33955835481495455, 1.7494161615202275, -0.15394117603733548], [-0.3419097544041847, -0.23992883666045373, -0.15587790880447727, 1.7292393175137022]], 'observation_offsets': [165.2279084503762, 157.76807691937614, 168.4235495099334, 225.33433430227353], 'transition_covariance': [[2515479496.145993, -401423541.70620924, 1409951418.1627903, 255932902.74454522], [-401423541.706214, 2744353887.676857, 1162316.2019491254, 1857251491.3987627], [1409951418.1628358, 1162316.2020361447, 543279068599.8229, -39399311190.5746], [255932902.74459982, 1857251491.398838, -39399311190.574585, 537826124257.5266]], 'transition_matrices': [[0.52163952865412, 0.47872618354122665, -0.0004322286766109684, 0.00017293351811531466], [0.5167436693545113, 0.48319044922845933, 7.765428142114672e-05, -0.00021518950285326355], [0.2091705950622469, 0.41051399729482796, 0.19341113299389256, 0.19562916616052917], [0.368592004009912, 0.22263632461118732, 0.20756792378812872, 0.20977025833570906]], 'transition_offsets': [592.5708159274, 583.3804671015271, 414.4187239098291, 562.166786712371]} echo_vm_5k={'initial_state_covariance': [[0.375, 0.0, 0.0, 0.0], [0.0, 0.375, 0.0, 0.0], [0.0, 0.0, 0.375, 0.0], [0.0, 0.0, 0.0, 0.375]], 'initial_state_mean': [15997944.198361743, 16029825.435899183, 17093077.26228404, 17524263.088803563], 'observation_covariance': [[36572556646.179054, 21816054953.37006, 31144379008.310543, 19651005729.823025], [21816054953.372543, 440428106325.20325, 41103447776.740585, 427146570672.51227], [31144379008.31037, 41103447776.74027, 3280009435458.6953, 458734528073.65686], [19651005729.82234, 427146570672.5109, 458734528073.6557, 3769493190697.773]], 'observation_matrices': [[1.0248853427592337, -0.031198859962501047, 0.001613706836380402, 0.004720209443291878], [-0.8604422900368718, 1.8583369609057172, -0.0022646214457040514, 0.004437933935378169], [-0.5814771409524866, 0.22228184387142846, 1.6259599749174072, -0.271594798325566], [-0.5862601003257453, 0.2598285939005791, -0.28286590143513024, 1.604087079832425]], 'observation_offsets': [1979.4518332096984, 1889.3380163762793, 2132.9112026744906, 1750.7759421584785], 'transition_covariance': [[6176492087.271547, 762254719.4171592, 4584288694.652873, 3044796192.4357214], [762254719.4185101, 173302376079.4761, 5261303152.757347, 167562483383.9925], [4584288694.651718, 5261303152.755746, 1056156956874.4131, -115859156952.07962], [3044796192.434162, 167562483383.9901, -115859156952.08018, 1225788436266.3086]], 'transition_matrices': [[0.9673912485796876, 0.03252962227543321, 0.0006756067792537124, -0.0006566638567164773], [0.9548761966068113, 0.03841774395880293, 0.00426067282319309, 0.002303362691861821], [0.6215040230859188, -0.2584476837756142, 0.3176491193420503, 0.3241682768126566], [0.6634028281470279, -0.33548335246018723, 0.3298144902195048, 0.3475836278392421]], 'transition_offsets': [1751.3049487348183, 1764.989515773476, 1986.8405778425586, 2232.830254345267]} #testKalman4D(echo_vm_5k) def testKalman(params=None): from pykalman import AdditiveUnscentedKalmanFilter,KalmanFilter train = db.subseries('train','long', offset=0) test = db.subseries('test','long', offset=0) measurements = numpy.asarray([(s['unusual_packet'],s['other_packet']) for s in (train+test)]) #kf = KalmanFilter(transition_matrices = [[1, 1], [0, 1]], observation_matrices = [[0.1, 0.5], [-0.3, 0.0]]) kf = KalmanFilter(n_dim_obs=2, n_dim_state=2, initial_state_mean=[quadsummary([s['unusual_packet'] for s in train]), quadsummary([s['other_packet'] for s in train])]) #kf = AdditiveUnscentedKalmanFilter(n_dim_obs=2, n_dim_state=2) if params == None: start=time.time() kf = kf.em(measurements[0:len(train)], n_iter=10, em_vars=('transition_matrices', 'observation_matrices', 'transition_offsets', 'observation_offsets', 'transition_covariance', 'observation_covariance', 'initial_state_covariance')) params = {'transition_matrices': kf.transition_matrices.tolist(), 'observation_matrices': kf.observation_matrices.tolist(), 'transition_offsets': kf.transition_offsets.tolist(), 'observation_offsets': kf.observation_offsets.tolist(), 'transition_covariance': kf.transition_covariance.tolist(), 'observation_covariance': kf.observation_covariance.tolist(), 'initial_state_mean': kf.initial_state_mean.tolist(), 'initial_state_covariance': kf.initial_state_covariance.tolist()} print("Learned Params:\n") import pprint pprint.pprint(params) print("pykalman em time: %f" % (time.time()-start)) #kf = KalmanFilter(n_dim_obs=2, n_dim_state=2, **params) num_obs=10000 for offset in range(50000,100000+num_obs,num_obs): start=time.time() kf = KalmanFilter(n_dim_obs=2, n_dim_state=2, **params) m = measurements[offset:offset+num_obs] (smoothed_state_means, smoothed_state_covariances) = kf.smooth(m) print("pykalman smooth time: %f" % (time.time()-start)) up = numpy.mean([m[0] for m in smoothed_state_means]) op = numpy.mean([m[1] for m in smoothed_state_means]) print("packet_rtt pykalman final:", smoothed_state_means[-1][0]-smoothed_state_means[-1][1]) print("packet_rtt pykalman mean:", up-op) print("packet_rtt mean:", numpy.mean([s[0]-s[1] for s in m])) five_iter = {'observation_offsets': [-54.53185823, -55.25219184], 'observation_covariance': [[ 1.15059170e+10, 4.36743765e+09], [ 4.36743765e+09, 1.19410313e+10]], 'initial_state_mean': [ 12939012.5625 , 12934563.71875], 'transition_covariance': [[ 2.98594543e+09, 6.86355073e+07], [ 6.86355073e+07, 3.21368699e+09]], 'initial_state_covariance': [[ 2.36836696e+09, 1.63195635e+09], [ 1.63195635e+09, 1.12452233e+09]], 'transition_offsets': [ 343.69740217, 338.5042467 ], 'observation_matrices': [[ 1.42539895, -0.4255261 ], [-0.46280375, 1.46295189]], 'transition_matrices': [[ 0.56151623, 0.4385931 ], [ 0.47309189, 0.52673508]]} ten_iter = {'initial_state_covariance': [[229936928.28125, 41172601.0], [41172601.0, 7372383.46875]], 'initial_state_mean': [12939012.5625, 12934563.71875], 'observation_covariance': [[11958914107.88334, 4761048283.066559], [4761048283.066557, 12388186543.42032]], 'observation_matrices': [[1.4258395826727792, -0.42598392357467674], [-0.4647443890462455, 1.4648767294384015]], 'observation_offsets': [165.409715349344, 157.96206130876212], 'transition_covariance': [[2515594742.7187943, -401728959.41375697], [-401728959.41375697, 2743831805.402682]], 'transition_matrices': [[0.521306461057975, 0.47879632652984583], [0.5167881285851763, 0.483006520280469]], 'transition_offsets': [592.4419187566978, 583.2272403965366]} #testKalman(ten_iter) def getTCPTSPrecision(): cursor = db.conn.cursor() query="""SELECT tcpts_mean FROM meta""" cursor.execute(query) row = cursor.fetchone() if row: return row[0] return None def tsFilteredHistogram(): tcpts_precision = getTCPTSPrecision() num_bins = 500 all = db.subseries('train','long')+db.subseries('test','long') diffs = [s['unusual_packet']-s['other_packet'] for s in all] ts0_diffs = [s['unusual_packet']-s['other_packet'] for s in all if s['unusual_tsval']-s['other_tsval'] == 0] ts1_diffs = [s['unusual_packet']-s['other_packet'] for s in all if abs(s['unusual_tsval']-s['other_tsval']) > 0] ts2_diffs = [s['unusual_packet']-s['other_packet'] for s in all if abs(round((s['unusual_tsval']-s['other_tsval'])/tcpts_precision)) <= 1.0] ts_mode = statistics.mode([s['unusual_tsval'] for s in all]+[s['other_tsval'] for s in all]) ts_diff_mode = statistics.mode([s['unusual_tsval']-s['other_tsval'] for s in all]) ts_common_mode = [s['unusual_packet']-s['other_packet'] for s in all if s['unusual_tsval']<=ts_mode and s['other_tsval']<=ts_mode] ts_common_diff_mode = [s['unusual_packet']-s['other_packet'] for s in all if s['unusual_tsval']-s['other_tsval']==ts_diff_mode] print('packet_rtt diff quadsummary: %f' % quadsummary(diffs)) print('packet_rtt tsval diff=0 quadsummary: %f' % quadsummary(ts0_diffs)) print('packet_rtt tsval diff>0 quadsummary: %f' % quadsummary(ts1_diffs)) print('packet_rtt tsval diff<=1 quadsummary: %f' % quadsummary(ts2_diffs)) print('packet_rtt tsval mode quadsummary: %f' % quadsummary(ts_common_mode)) print(len(diffs), len(ts0_diffs)+len(ts1_diffs)) diffs.sort() cut_off_low = diffs[int(len(diffs)*0.005)] cut_off_high = diffs[int(len(diffs)*0.995)] plt.clf() # the histogram of the data n, bins, patches = plt.hist(diffs, num_bins, normed=0, color='black', histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high), label='all') n, bins, patches = plt.hist(ts0_diffs, num_bins, normed=0, color='blue', histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high), label='tsval diff=0') n, bins, patches = plt.hist(ts1_diffs, num_bins, normed=0, color='red', histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high), label='tsval diff>0') n, bins, patches = plt.hist(ts2_diffs, num_bins, normed=0, color='orange', histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high), label='tsval diff<=1') #n, bins, patches = plt.hist(ts_common_mode, num_bins, normed=0, color='green', histtype='step', alpha=0.8, # range=(cut_off_low,cut_off_high), label='tsval common mode') n, bins, patches = plt.hist(ts_common_diff_mode, num_bins, normed=0, color='green', histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high), label='tsval common diff mode') plt.xlabel('RTT Difference') plt.ylabel('Probability') plt.title(r'Histogram - distribution of differences by tsval') # Tweak spacing to prevent clipping of ylabel plt.subplots_adjust(left=0.15) plt.legend() plt.show() #plt.savefig('paper/graphs/dists-vs-dist-of-diffs2.svg') #tsFilteredHistogram() #all_data = longs+shorts #all_data.sort() #cut_off_low = all_data[0] #cut_off_high = all_data[int(len(all_data)*0.997)] def plotSingleProbe(probe_id=None): if probe_id == None: cursor = db.conn.cursor() query="""SELECT probe_id FROM analysis WHERE suspect='' ORDER BY probe_id DESC limit 1 OFFSET 10""" cursor.execute(query) probe_id = cursor.fetchone()[0] cursor = db.conn.cursor() query="""SELECT observed,payload_len FROM packets WHERE probe_id=? AND sent=1""" cursor.execute(query, (probe_id,)) pkts = cursor.fetchall() sent_payload = [row[0] for row in pkts if row[1] != 0] sent_other = [row[0] for row in pkts if row[1] == 0] query="""SELECT observed,payload_len FROM packets WHERE probe_id=? AND sent=0""" cursor.execute(query, (probe_id,)) pkts = cursor.fetchall() rcvd_payload = [row[0] for row in pkts if row[1] != 0] rcvd_other = [row[0] for row in pkts if row[1] == 0] #query="""SELECT reported,time_of_day FROM probes WHERE id=?""" #cursor.execute(query, (probe_id,)) #reported,tod = cursor.fetchone() #userspace_times = [sent_times[0]-reported/3.0, sent_times[0]+reported] print("single probe counts:",len(sent_payload),len(sent_other),len(rcvd_payload),len(rcvd_other)) plt.clf() plt.title("Single HTTP Request - Packet Times") sp = plt.eventplot(sent_payload, colors=('red',), lineoffsets=8, linewidths=2, alpha=0.6,label='sent') so = plt.eventplot(sent_other, colors=('red',), lineoffsets=6, linewidths=2, alpha=0.6,label='sent') rp = plt.eventplot(rcvd_payload, colors=('blue',), lineoffsets=4, linewidths=2, alpha=0.6,label='received') ro = plt.eventplot(rcvd_other, colors=('blue',), lineoffsets=2, linewidths=2, alpha=0.6,label='received') #plt.legend((s,r), ('sent','received')) #plt.savefig('../img/http-packet-times.svg') plt.show() #plotSingleProbe() def graphTestResults(): plt.clf() plt.title("Test Results") plt.xlabel('sample size') plt.ylabel('error rate') legend = [] colors = ['red','blue','green','purple','orange','black','brown'] color_id = 0 cursor = db.conn.cursor() query = """ SELECT classifier FROM classifier_results GROUP BY classifier ORDER BY classifier; """ cursor.execute(query) classifiers = [] for c in cursor: classifiers.append(c[0]) max_obs = 0 for classifier in classifiers: query=""" SELECT params FROM classifier_results WHERE trial_type='test' AND classifier=:classifier AND (false_positives+false_negatives)/2.0 < 5.0 ORDER BY num_observations,(false_positives+false_negatives) LIMIT 1 """ cursor.execute(query, {'classifier':classifier}) row = cursor.fetchone() if row == None: query=""" SELECT params FROM classifier_results WHERE trial_type='test' and classifier=:classifier ORDER BY (false_positives+false_negatives),num_observations LIMIT 1 """ cursor.execute(query, {'classifier':classifier}) row = cursor.fetchone() if row == None: sys.stderr.write("WARN: couldn't find test results for classifier '%s'.\n" % classifier) continue best_params = row[0] query=""" SELECT num_observations,(false_positives+false_negatives)/2.0 FROM classifier_results WHERE trial_type='test' AND classifier=:classifier AND params=:params ORDER BY num_observations """ cursor.execute(query, {'classifier':classifier,'params':best_params}) num_obs = [] performance = [] for row in cursor: max_obs = max(max_obs, row[0]) num_obs.append(row[0]) performance.append(row[1]) #print(num_obs,performance) path = plt.scatter(num_obs, performance, color=colors[color_id], s=4, alpha=0.8, linewidths=3.0) plt.plot(num_obs, performance, color=colors[color_id], alpha=0.8) legend.append((classifier,path)) color_id = (color_id+1) % len(colors) plt.legend([l[1] for l in legend], [l[0] for l in legend], scatterpoints=1, fontsize='xx-small') plt.plot([0, max_obs], [5.0, 5.0], "k--") plt.show() graphTestResults() sys.exit(0) plt.clf() plt.title("Packet RTT over time") plt.xlabel('Time of Day') plt.ylabel('RTT') s = plt.scatter([t for t,rtt in short_overtime], [rtt for t,rtt in short_overtime], s=1, color='red', alpha=0.6) l = plt.scatter([t for t,rtt in long_overtime], [rtt for t,rtt in long_overtime], s=1, color='blue', alpha=0.6) d = plt.scatter([t for t,rtt in diff_overtime], [rtt for t,rtt in diff_overtime], s=1, color='purple', alpha=0.6) plt.legend((s,l,d), ('short','long','difference'), scatterpoints=1) #plt.savefig('paper/figures/comcast-powerboost1.png') plt.show() plt.clf() plt.title("Simple HTTP Request") plt.xlabel('Time of Day') plt.ylabel('') s = plt.scatter(sent_times, [2]*len(sent_times), s=3, color='red', alpha=0.9) r = plt.scatter(rcvd_times, [1]*len(rcvd_times), s=3, color='blue', alpha=0.9) plt.legend((s,r), ('sent','received'), scatterpoints=1) plt.show() sys.exit(0) short_overtime,long_overtime,diff_overtime = None,None,None num_bins = 300 reported_diffs.sort() cut_off_low = reported_diffs[int(len(diffs)*0.003)] cut_off_high = reported_diffs[int(len(diffs)*0.997)] plt.clf() # the histogram of the data n, bins, patches = plt.hist(reported_diffs, num_bins, normed=1, color='black', histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high)) plt.xlabel('RTT Difference') plt.ylabel('Probability') plt.title(r'Histogram - distribution of differences') # Tweak spacing to prevent clipping of ylabel plt.subplots_adjust(left=0.15) #plt.legend() plt.show() #plt.savefig('paper/graphs/dists-vs-dist-of-diffs2.svg') num_bins = 300 diffs.sort() cut_off_low = diffs[int(len(diffs)*0.003)] cut_off_high = diffs[int(len(diffs)*0.997)] plt.clf() # the histogram of the data n, bins, patches = plt.hist(diffs, num_bins, normed=1, color='purple', histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high)) plt.xlabel('RTT Difference') plt.ylabel('Probability') plt.title(r'Histogram - distribution of differences') # Tweak spacing to prevent clipping of ylabel plt.subplots_adjust(left=0.15) #plt.legend() plt.show() #plt.savefig('paper/graphs/dists-vs-dist-of-diffs2.svg') sys.exit(0) num_bins = 150 # the histogram of the data n, bins, patches = plt.hist((shorts,longs), num_bins, normed=1, label=['short', 'long'], color=['red','blue'], histtype='step', alpha=0.8, range=(cut_off_low,cut_off_high)) #n, bins, patches = plt.hist(shorts2+longs2, num_bins, normed=1, facecolor='blue', histtype='step', alpha=0.3) # add a 'best fit' line #y = mlab.normpdf(bins, mu, sigma) #plt.plot(bins, y, 'r--') plt.xlabel('packet_rtt') plt.ylabel('Probability') plt.title(r'Histogram - RTT short and long') # Tweak spacing to prevent clipping of ylabel plt.subplots_adjust(left=0.15) plt.legend() #plt.show() plt.savefig('paper/figures/comcast-powerboost2.svg') num_trials = 200 subsample_sizes = (50,150,300,500,700,1000,2000,3000,5000,7000,10000,15000,20000) estimator = functools.partial(boxTest, 0.07, 0.08) performance = [] for subsample_size in subsample_sizes: estimates = bootstrap(derived, subsample_size, num_trials, estimator) performance.append(100.0*len([e for e in estimates if e == 1])/num_trials) null_performance = [] for subsample_size in subsample_sizes: null_estimates = bootstrap(null_derived, subsample_size, num_trials, estimator) null_performance.append(100.0*len([e for e in null_estimates if e == 0])/num_trials) plt.clf() plt.title("boxTest bootstrap") plt.xlabel('sample size') plt.ylabel('performance') plt.scatter(subsample_sizes, performance, s=2, color='red', alpha=0.6) plt.scatter(subsample_sizes, null_performance, s=2, color='blue', alpha=0.6) plt.show() subsample_sizes = (50,150,300,400,500,700,1000,2000,3000,4000,5000,7000,10000) estimator = diffMedian performance = [] for subsample_size in subsample_sizes: estimates = bootstrap(derived, subsample_size, num_trials, estimator) performance.append(100.0*len([e for e in estimates if e > expected_mean*0.9 and e < expected_mean*1.1])/num_trials) plt.clf() plt.title("diff median bootstrap") plt.xlabel('sample size') plt.ylabel('performance') plt.scatter(subsample_sizes, performance, s=1, color='red', alpha=0.6) plt.show() subsample_sizes = (50,150,300,400,500,700,1000,2000,3000,4000,5000,7000,10000) weight_funcs = (linearWeights, prunedWeights) for wf in weight_funcs: estimator = functools.partial(estimateMean, hypotenuse, wf, 0.40) performance = [] for subsample_size in subsample_sizes: estimates = bootstrap(derived, subsample_size, num_trials, estimator) performance.append(100.0*len([e for e in estimates if e > expected_mean*0.9 and e < expected_mean*1.1])/num_trials) plt.clf() plt.title(repr(wf)) plt.xlabel('sample size') plt.ylabel('performance') plt.scatter(subsample_sizes, performance, s=1, color='red', alpha=0.6) plt.show() num_bins = 300 # the histogram of the data n, bins, patches = plt.hist((tsshorts,tslongs), num_bins, normed=1, label=['short', 'long'], color=['red','blue'], histtype='step', alpha=0.8) #n, bins, patches = plt.hist(shorts2+longs2, num_bins, normed=1, facecolor='blue', histtype='step', alpha=0.3) # add a 'best fit' line #y = mlab.normpdf(bins, mu, sigma) #plt.plot(bins, y, 'r--') plt.xlabel('packet_rtt') plt.ylabel('Probability') plt.title(r'Histogram - tsval_rtt short vs long') # Tweak spacing to prevent clipping of ylabel plt.subplots_adjust(left=0.15) plt.legend() plt.show() #### #trust_methods = [min,max,sum,difference,product] trust_methods = [sum,product,hypotenuse] colors = ['red','blue','green','purple','orange','black'] weight_methods = [prunedWeights, linearWeights] alphas = [i/100.0 for i in range(0,100,2)] plt.clf() plt.title(r'Trust Method Comparison - Linear') plt.xlabel('Alpha') plt.ylabel('Mean error') paths = [] for tm in trust_methods: trust = trustValues(derived, tm) series = [] for alpha in alphas: weights = linearWeights(derived, trust, alpha) series.append(weightedMean(derived, weights) - expected_mean) paths.append(plt.scatter(alphas, series, s=1, color=colors[len(paths)],alpha=0.6)) plt.legend(paths, [repr(tm) for tm in trust_methods], scatterpoints=1) plt.show() plt.clf() plt.title(r'Trust Method Comparison - Pruned') plt.xlabel('Alpha') plt.ylabel('Mean error') paths = [] for tm in trust_methods: trust = trustValues(derived, tm) series = [] for alpha in alphas: weights = prunedWeights(derived, trust, alpha) series.append(weightedMean(derived, weights) - expected_mean) paths.append(plt.scatter(alphas, series, s=1, color=colors[len(paths)],alpha=0.6)) plt.legend(paths, [repr(tm) for tm in trust_methods], scatterpoints=1) plt.show() sys.exit(0) plt.clf() plt.title(r'Trust Method Comparison - Inverted') plt.xlabel('Alpha') plt.ylabel('Mean error') paths = [] for tm in trust_methods: trust = trustValues(derived, tm) series = [] for alpha in alphas: weights = invertedWeights(derived, trust, alpha) series.append(weightedMean(derived, weights) - expected_mean) paths.append(plt.scatter(alphas, series, s=1, color=colors[len(paths)],alpha=0.6)) plt.legend(paths, [repr(tm) for tm in trust_methods], scatterpoints=1) plt.show() plt.clf() plt.title(r'Trust Method Comparison - Arctangent') plt.xlabel('Alpha') plt.ylabel('Mean error') paths = [] for tm in trust_methods: trust = trustValues(derived, tm) series = [] for alpha in alphas: weights = arctanWeights(derived, trust, alpha) series.append(weightedMean(derived, weights) - expected_mean) paths.append(plt.scatter(alphas, series, s=1, color=colors[len(paths)],alpha=0.6)) plt.legend(paths, [repr(tm) for tm in trust_methods], scatterpoints=1) plt.show()