def locations(path=None):
_, y = get_train_data(path)
if len(y) == 0: # pragma: no cover
msg = "No location samples available. First learn a location, e.g. with `whereami learn -l kitchen`."
print(msg)
else:
occurrences = Counter(y)
for key, value in occurrences.items():
print("{}: {}".format(key, value))
def train_model():
model_file = get_model_file()
X, y = get_train_data()
# fantastic: because using "quality" rather than "rssi", we expect values 0-150
# 0 essentially indicates no connection
# 150 is something like best possible connection
# Not observing a wifi will mean a value of 0, which is the perfect default.
lp = get_pipeline()
lp.fit(X, y)
with open(model_file, "wb") as f:
pickle.dump(lp, f)
return lp
def train_model():
model_file = get_model_file()
X, y = get_train_data()
# fantastic: because using "quality" rather than "rssi", we expect values 0-150
# 0 essentially indicates no connection
# 150 is something like best possible connection
# Not observing a wifi will mean a value of 0, which is the perfect default.
lp = get_pipeline()
lp.fit(X, y)
with open(model_file, "wb") as f:
pickle.dump(lp, f)
return lp
def crossval(clf=None, X=None, y=None, folds=10, n=5):
if X is None or y is None:
X, y = get_train_data()
clf = clf or get_model()
tot = 0
print("KFold folds={}, running {} times".format(folds, n))
for i in range(n):
res = cross_val_score(clf, X, y, cv=folds).mean()
tot += res
print("{}/{}: {}".format(i + 1, n, res))
print("-------- total --------")
print(tot / n)
return tot / n
def train_model(path=None):
model_file = get_model_file(path)
X, y = get_train_data(path)
if len(X) == 0:
raise ValueError("No wifi access points have been found during training")
# fantastic: because using "quality" rather than "rssi", we expect values 0-150
# 0 essentially indicates no connection
# 150 is something like best possible connection
# Not observing a wifi will mean a value of 0, which is the perfect default.
lp = get_pipeline()
lp.fit(X, y)
with open(model_file, "wb") as f:
pickle.dump(lp, f)
return lp
def crossval(clf=None, X=None, y=None, folds=10, n=5, path=None):
if X is None or y is None:
X, y = get_train_data(path)
if len(X) < folds:
raise ValueError('There are not enough samples ({}). Need at least {}.'.format(len(X), folds))
clf = clf or get_model(path)
tot = 0
print("KFold folds={}, running {} times".format(folds, n))
for i in range(n):
res = cross_val_score(clf, X, y, cv=folds).mean()
tot += res
print("{}/{}: {}".format(i + 1, n, res))
print("-------- total --------")
print(tot / n)
return tot / n
def crossval():
X, y = get_train_data()
lp = get_model()
print(cross_validate_model(lp, X, y))