-
Notifications
You must be signed in to change notification settings - Fork 140
/
analyze_stats.py
57 lines (49 loc) · 1.74 KB
/
analyze_stats.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
import numpy as np
import cPickle
import matplotlib.pyplot as plt
def calc_vec_energy(vec):
isquared = np.power(vec[0],2.0)
qsquared = np.power(vec[1], 2.0)
inst_energy = np.sqrt(isquared+qsquared)
return sum(inst_energy)
def calc_mod_energies(ds):
for modulation, snr in ds:
avg_energy = 0
nvectors = ds[(modulation,snr)].shape[0]
for vec in ds[(modulation, snr)]:
avg_energy += calc_vec_energy(vec)
avg_energy /= nvectors
print "%s at %i has %i vectors avg energy of %2.1f" % (modulation, snr, nvectors, avg_energy)
def calc_mod_bias(ds):
for modulation, snr in ds:
avg_bias_re = 0
avg_bias_im = 0
nvectors = ds[(modulation,snr)].shape[0]
for vec in ds[(modulation, snr)]:
avg_bias_re += (np.mean(vec[0]))
avg_bias_im += (np.mean(vec[1]))
#avg_bias_re /= nvectors
#avg_bias_im /= nvectors
print "%s at %i has %i vectors avg bias of %2.1f + %2.1f j" % (modulation, snr, nvectors, avg_bias_re, avg_bias_im)
def calc_mod_stddev(ds):
for modulation, snr in ds:
avg_stddev = 0
nvectors = ds[(modulation,snr)].shape[0]
for vec in ds[(modulation, snr)]:
avg_stddev += np.abs(np.std(vec[0]+1j*vec[1]))
#avg_stddev /= nvectors
print "%s at %i has %i vectors avg stddev of %2.1f" % (modulation, snr, nvectors, avg_stddev)
def open_ds(location="X_4_dict.dat"):
f = open(location)
ds = cPickle.load(f)
return ds
def main():
ds = open_ds()
#plt.plot(ds[('BPSK', 12)][25][0][:])
#plt.plot(ds[('BPSK', 12)][25][1][:])
#plt.show()
#calc_mod_energies(ds)
#calc_mod_stddev(ds)
calc_mod_bias(ds)
if __name__ == "__main__":
main()