-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathplot_n1_latency.py
236 lines (209 loc) · 8.89 KB
/
plot_n1_latency.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
import sys
import os
import matplotlib.pyplot as plt
import math
#farr = [10, 15, 20, 25, 30, 33, 38, 45, 55, 67, 80, 100]
farr = [10, 15, 19, 21, 22, 23, 25, 30, 33, 38, 45, 67, 80, 100] #Smallc1
#farr = [9, 12, 15, 17, 19, 21, 23, 30, 45, 67, 80, 100] #Largec1
#farr = [10, 15, 17, 20, 23, 26, 28, 30, 32, 35, 40, 60, 80, 100] # Smallc1 2c
# farr = [12, 20, 28, 36, 44, 52, 60, 70, 80, 90, 100, 130]
farr = [9, 16, 23, 24, 25, 30, 55, 80]
pre = ''
# argv1 : name of file with actual freqs
# argv2 : prefix of log files
if __name__ == '__main__':
pre = sys.argv[1]
cpp = int(sys.argv[2])
need_actual_freq = (int(sys.argv[7]) == 1) # we dont need new_freq for RTC and DynamicAlgo.
farr = farr if need_actual_freq else [15]
fname = sys.argv[8]
mean_lats = {}
med_lats = {}
new_farrs = {}
# ci_264kb = {"_Newc1" : 20, "_c1" : 26, "_Medc1" : 16}
# ci_1mb = {"" : 16, "_Newc1" : 22, "_c1" : 34} #
# newc = {"_c1A" : 22, "_c1BA" : 30, "_c1BB" : 30} # "_c1A" : 22, "_c1AB" : 22, : python
# newc = {"_c1" : 18, "_c1B" : 22, "_c1D" : 38} # cpp
c1 = int(sys.argv[5])
# newc = {"" : int(sys.argv[5])}
# if sys.argv[4] == "1mb":
# ci = ci_1mb
# elif sys.argv[4] == "264kb":
# ci = ci_264kb
# else:
# ci = newc
t=int(sys.argv[6])
# for c1 in ci.keys():
# print "Starting ", c1
# if c1 == "" and sys.argv[4] == "1mb":
# farr[1] = 10
# # else:
# # farr[1] = 12
# # if c1 == "_c1A" and sys.argv[4] == "new":
# # farr[-2] = 115
# # else:
# # farr[-2] = 120
ind = {}
for i in range(len(farr)):
ind[farr[i]] = i
perc_lat = [0.0 for x in farr]
med_lat = [0.0 for x in farr]
mean_lat = [0.0 for x in farr]
mean_perc_lat = [0.0 for x in farr]
real_perc_lat = [0.0 for x in farr]
real_med_lat = [0.0 for x in farr]
real_mean_lat = [0.0 for x in farr]
td_perc_lat = [0.0 for x in farr]
td_med_lat = [0.0 for x in farr]
td_mean_lat = [0.0 for x in farr]
new_farr = [0.0 for x in farr]
print farr
pre1 = pre
if need_actual_freq:
with open(pre1 + "_actual_freq.txt", 'r') as af:
afl = af.readlines()
for l in afl:
ls = l.split(' ')
freq = int(ls[0])
if int(ls[1]) == t and freq in farr:
new_farr[ind[freq]] = float(ls[-1][:-1])
else:
new_farr = farr
runs = [1,2,3,4,5,6,7,8,9,10]
for f in farr:
for r in runs:
with open('%s_preprocess_node_%i.%i.%i.out'%(pre1, r, f, t), 'r') as fil:
print "Reading for ", f, t, pre1, r
for l in fil.readlines():
larr = l.split(' ')
if (cpp == 1):
# c1n_latency if roscpp files :
if 'c1n_latency' in l:
if ":," in l:
pl= float(larr[2][:-1])
medl = float(larr[3][:-1])
meanl = float(larr[4][:-1])
else:
pl = float(larr[6][:-1])
medl = float(larr[7][:-1])
meanl = float(larr[8][:-1])
elif "Latency w.r.t. TDNode" in l:
tdpl = float(larr[12])
tdmedl = float(larr[13])
tdmeanl = float(larr[14])
else:
if 'latency of msg arrival at N1' in l:
perc_lat[ind[f]] = float(larr[14][:-2])
med_lat[ind[f]] = float(larr[13][:-1])
mean_lat[ind[f]] = float(larr[12][:-1])
perc_lat[ind[f]] += pl
med_lat[ind[f]] += medl
mean_lat[ind[f]] += meanl
td_perc_lat[ind[f]] += tdpl
td_med_lat[ind[f]] += tdmedl
td_mean_lat[ind[f]] += tdmeanl
# print perc_lat, med_lat, td_mean_lat
#average over 10 runs:
perc_lat[ind[f]] /= len(runs)
med_lat[ind[f]] /= len(runs)
mean_lat[ind[f]] /= len(runs)
td_perc_lat[ind[f]] /= len(runs)
td_med_lat[ind[f]] /= len(runs)
td_mean_lat[ind[f]] /= len(runs)
# if (cpp == 0):
# with open(pre1 + '_' + ss + 'preprocess_lat_' + str(f) + str(t) + '.txt', 'r') as ff:
# arr = [x.split(' ') for x in ff.readlines()[:-1]]
# lat_arr = [float(x[2][:-1]) for x in arr]
# ll = len(lat_arr)
# lat_arr = sorted(lat_arr)
# if perc_lat[ind[f]] == 0:
# perc_lat[ind[f]] = lat_arr[(95*ll)/100]
# med_lat[ind[f]] = lat_arr[ll/2]
# mean_lat[ind[f]] = sum(lat_arr)/ll
# real_recv_times = {}
# for x in arr:
# real_recv_times[int(x[0])] = float(x[1])
# # print real_recv_times[3000], "real recv time at N1 for msg id 5"
# # for real lat : read real send time from gz logs :
# with open('/home/aditi/catkin_ws/Apr_Cam_RT_Logs/' + pre1 + '_CamLogs_' + str(f) + '.out', 'r') as fil:
# arr = [x.split(' ') for x in fil.readlines()[:-1]]
# real_msg_ts = [0.0 for x in arr]
# for x in arr:
# real_msg_ts[int(x[0])] = float(x[2][:-1])
# print real_msg_ts[5], real_msg_ts[17]
# real_lat_arr = []
# for k in sorted(real_recv_times.keys()):
# real_lat_arr.append(real_recv_times[k] - real_msg_ts[k])
# if k%500 == 3:
# print real_lat_arr[-1], k
# real_lat_arr.sort()
# rl = len(real_lat_arr)
# real_perc_lat[ind[f]] = real_lat_arr[(rl*95)/100]
# real_med_lat[ind[f]] = real_lat_arr[(rl)/2]
# real_mean_lat[ind[f]] = sum(real_lat_arr)/rl
# mean_perc_lat[ind[f]] = perc_lat[ind[f]] + mean_lat[ind[f]]
print new_farr
print perc_lat, med_lat, mean_lat
# write to file!
with open(fname, 'a') as f1:
f1.write('%i %i 10RunAvg N1Latency Tail, Med, Mean : %f %f %f #\n'%(f, t, perc_lat[ind[f]], med_lat[ind[f]], mean_lat[ind[f]]))
f1.write('%i %i 10RunAvg N1Latency w.r.t. TDNode Tail, Med, Mean : %f %f %f #\n'%(f, t, td_perc_lat[ind[f]], td_med_lat[ind[f]], td_mean_lat[ind[f]]))
print td_perc_lat, td_med_lat, td_mean_lat
x = 5
s = 'roscpp' if (cpp == 1) else 'rospy'
plt.plot(new_farr, perc_lat, 'ro-', label='9%iile'%x)
plt.plot(new_farr, mean_lat, 'b.--', label='mean')
plt.plot(new_farr, med_lat, 'g*:', label='Median')
plt.title('Mean, 9%iile Latency (gz->%s) c1=%dms, %s'%(x, s, c1, pre1))
plt.xlabel('Publisher Frequency')
plt.ylabel('Latency')
plt.ylim(0, 0.095)
plt.legend()
plt.show()
plt.plot(new_farr, td_perc_lat, 'ro-', label='9%iile'%x)
plt.plot(new_farr, td_mean_lat, 'b.--', label='mean')
plt.plot(new_farr, td_med_lat, 'g*:', label='Median')
plt.title('Mean, 9%iile Latency w.r.t. TDNode (gz->%s) c1=%dms, %s'%(x, s, c1, pre1))
plt.xlabel('Publisher Frequency')
plt.ylabel('Latency wrt TD')
plt.legend()
plt.show()
plt.plot(new_farr, real_perc_lat, 'ro-', label='9%iile'%x)
plt.plot(new_farr, real_mean_lat, 'b.--', label='mean')
plt.plot(new_farr, real_med_lat, 'g*:', label='Median')
plt.title('Mean, 9%iile RealTime Latency (gz->%s) c1=%dms, %s'%(x, s, c1, pre1))
plt.xlabel('Publisher Frequency')
plt.ylabel('RT Latency')
plt.legend()
plt.show()
mean_lats[c1] = mean_lat
med_lats[c1] = med_lat
new_farrs[c1] = new_farr
# for x in ci.keys():
# plt.plot(new_farrs[x], mean_lats[x], '*:', label='Mean for %dms, %s'%(ci[x], pre))
# plt.title("Mean latency at " + sys.argv[3])
# plt.xlabel('Publisher Frequency')
# plt.ylabel('Mean latency')
# plt.legend()
# plt.show()
# for x in ci.keys():
# plt.plot(new_farrs[x], med_lats[x], '*:', label='Median for %dms, %s'%(ci[x], pre))
# plt.title("Median latency at " + sys.argv[3])
# plt.xlabel('Publisher Frequency')
# plt.ylabel('Median latency')
# plt.legend()
# plt.show()
# latency w.r.t. time :
# with open(sys.argv[4], 'r') as f:
# larr = f.readlines()
# num_elem = len(larr) - 2
# iarr = [i for i in range(num_elem)]
# lat_arr = []
# for l in larr[2:]:
# lat_arr.append(float(l[:-1]))
# plt.plot(iarr, lat_arr, 'o-', label='Msg Latency')
# plt.title('Latency w.r.t. Time at 60Hz (Actual : 38Hz)')
# plt.xlabel('Received Message Number')
# plt.ylabel('Latency')
# plt.legend()
# plt.show()