def test_process(self):
MAJOR_AP_COUNT = 17
BAD_SIGNAL = -100
areaf = open("sample_data/wrm/wrm.json.dat")
area_json_list = areaf.readlines()
areaf.close()
area_set = {}
ap_set = {}
def compare(x, y):
if x['rssi'] < y['rssi']:
return 1
elif x['rssi'] == y['rssi']:
return 0
else:
return -1
for aobj in area_json_list:
area = json.loads(aobj)
label = area["areaID"]
aplist = area["apList"]
aplist.sort(compare)
for ap in aplist[:MAJOR_AP_COUNT]:
ap_set[ap['bssid']] = 1
area_set[label] = area_set.get(label, [])
area_set[label].append(aplist[:MAJOR_AP_COUNT])
ap_vector_column = ap_set.keys()
train_mat = []
train_label = []
test_mat = []
test_label = []
count = 0
for label in area_set:
for aps in area_set[label]:
ap_vector = tile(BAD_SIGNAL, len(ap_vector_column))
for ap in aps:
ap_vector[ap_vector_column.index(ap['bssid'])] = ap['rssi']
count += 1
if count % 10 == 0:
test_label.append(label)
test_mat.append(ap_vector)
else:
train_label.append(label)
train_mat.append(ap_vector)
gnb = GaussianNaiveBayes(train_mat, train_label)
gnb.fit()
error_rate = autotest.eval_predict(gnb, test_mat, test_label,
self.logging)
self.tlog("rssi predict (with GaussianNaiveBayes) error rate : " +
str(error_rate))
def test_process(self):
MAJOR_AP_COUNT = 17
BAD_SIGNAL = -100
areaf = open("sample_data/wrm/wrm.json.dat")
area_json_list = areaf.readlines()
areaf.close()
area_set = {}
ap_set = {}
def compare(x,y):
if x['rssi'] < y['rssi']:
return 1
elif x['rssi'] == y['rssi']:
return 0
else:
return -1
for aobj in area_json_list:
area = json.loads(aobj)
label = area["areaID"]
aplist = area["apList"]
aplist.sort(compare)
for ap in aplist[:MAJOR_AP_COUNT]:
ap_set[ap['bssid']] = 1
area_set[label] = area_set.get(label,[])
area_set[label].append(aplist[:MAJOR_AP_COUNT])
ap_vector_column = ap_set.keys()
train_mat = []
train_label = []
test_mat = []
test_label = []
count = 0;
for label in area_set:
for aps in area_set[label]:
ap_vector = tile(BAD_SIGNAL, len(ap_vector_column))
for ap in aps:
ap_vector[ap_vector_column.index(ap['bssid'])] = ap['rssi']
count += 1
if count % 10 == 0:
test_label.append(label)
test_mat.append(ap_vector)
else :
train_label.append(label)
train_mat.append(ap_vector)
gnb = GaussianNaiveBayes(train_mat,train_label)
gnb.fit()
error_rate = autotest.eval_predict(gnb, test_mat, test_label, self.logging)
self.tlog("rssi predict (with GaussianNaiveBayes) error rate : " + str(error_rate))
def test_process(self):
train_mat = [\
[-65,-55,-42],[-20,-59,-71],[-43,-49,-69],\
[-61,-30,-74],[-79,-81,-40],[-71,-57,-24],\
[-67,-19,-58],[-57,-73,-83],[-68,-74,-59],\
[-80,-85,-79]
]
train_label = ['B','A','A','A','B','B','A','C','C','C']
test_mat = [\
[-45,-47,-74],[-77,-69,-25],[-64,-71,-59],\
[-85,-85,-25],[-85,-85,-85]
]
test_label = ['A','B','C','B','C']
gnb = GaussianNaiveBayes(train_mat,train_label)
gnb.fit()
error_rate = autotest.eval_predict(gnb, test_mat, test_label, self.logging)
self.tlog("strength of signal predict error rate : " + str(error_rate))
def test_process(self):
train_mat = [\
[-65,-55,-42],[-20,-59,-71],[-43,-49,-69],\
[-61,-30,-74],[-79,-81,-40],[-71,-57,-24],\
[-67,-19,-58],[-57,-73,-83],[-68,-74,-59],\
[-80,-85,-79]
]
train_label = ['B', 'A', 'A', 'A', 'B', 'B', 'A', 'C', 'C', 'C']
test_mat = [\
[-45,-47,-74],[-77,-69,-25],[-64,-71,-59],\
[-85,-85,-25],[-85,-85,-85]
]
test_label = ['A', 'B', 'C', 'B', 'C']
gnb = GaussianNaiveBayes(train_mat, train_label)
gnb.fit()
error_rate = autotest.eval_predict(gnb, test_mat, test_label,
self.logging)
self.tlog("strength of signal predict error rate : " + str(error_rate))
