def main(tseries_fpath, train_fpath, test_fpath, assign_fpath, out_folder):
Xtrain = ioutil.load_series(tseries_fpath, train_fpath)
Xtest = ioutil.load_series(tseries_fpath, test_fpath)
y_train = np.genfromtxt(assign_fpath, dtype='i')
max_pts = Xtest.shape[1]
for num_pts in [1, 25, 50, 75]:
probs = fit(Xtrain, y_train, Xtest, num_pts)
probs_fpath = os.path.join(out_folder, 'probs-%d-pts.dat' % num_pts)
np.savetxt(probs_fpath, probs)
def main(tseries_fpath, train_fpath, test_fpath, assign_fpath, out_folder):
Xtrain = ioutil.load_series(tseries_fpath, train_fpath)
Xtest = ioutil.load_series(tseries_fpath, test_fpath)
y_train = np.genfromtxt(assign_fpath, dtype='i')
max_pts = Xtest.shape[1]
for num_pts in [1, 25, 50, 75]:
probs = fit(Xtrain, y_train, Xtest, num_pts)
probs_fpath = os.path.join(out_folder, 'probs-%d-pts.dat' % num_pts)
np.savetxt(probs_fpath, probs)
def run_fold(folder, tseries_fpath, min_pts, thetas, out_folder):
try:
os.makedirs(out_folder)
except:
pass
train_fpath = os.path.join(folder, "train.dat")
cents_fpath = os.path.join(folder, "ksc", "cents.dat")
assign_fpath = os.path.join(folder, "ksc", "assign.dat")
probs_folder = os.path.join(folder, "probs")
X = ioutil.load_series(tseries_fpath, train_fpath)
train_idx = np.loadtxt(train_fpath, dtype="bool")
y_true = np.loadtxt(assign_fpath)
num_series = X.shape[0]
max_pts = X.shape[1]
# Since we prune the first 100 lines of X we need to read other info
peak_days = []
sum_views = []
with open(tseries_fpath) as tseries_file:
for i, line in enumerate(tseries_file):
if train_idx[i]:
x = np.array([int(v) for v in line.split()[1:]])
peak_days.append(x.argmax())
sum_views.append(x.sum())
peak_days = np.array(peak_days)
sum_views = np.array(sum_views)
y_pred, best_by, confs, all_confs = pred(probs_folder, num_series, max_pts, min_pts, thetas)
save_results(X, peak_days, sum_views, min_pts, thetas, best_by, all_confs, y_true, y_pred, confs, out_folder)
def run_fold(folder, tseries_fpath, min_pts, thetas, out_folder, gamma_max):
try:
os.makedirs(out_folder)
except:
pass
test_fpath = os.path.join(folder, 'test.dat')
cents_fpath = os.path.join(folder, 'ksc', 'cents.dat')
assign_fpath = os.path.join(folder, 'ksc', 'test_assign.dat')
probs_folder = os.path.join(folder, 'probs-test')
X = ioutil.load_series(tseries_fpath, test_fpath)
test_idx = np.loadtxt(test_fpath, dtype='bool')
y_true = np.loadtxt(assign_fpath)
num_series = X.shape[0]
max_pts = gamma_max
peak_days = []
sum_views = []
with open(tseries_fpath) as tseries_file:
for i, line in enumerate(tseries_file):
if test_idx[i]:
x = np.array([int(v) for v in line.split()[1:]])
peak_days.append(x.argmax())
sum_views.append(x.sum())
peak_days = np.array(peak_days)
sum_views = np.array(sum_views)
y_pred, best_by, confs, all_confs = \
pred(probs_folder, num_series, max_pts, min_pts, thetas)
save_results(X, peak_days, sum_views, min_pts, thetas, best_by, all_confs,
y_true, y_pred, confs, out_folder)
def run_fold(folder, tseries_fpath, min_pts, thetas, out_folder, gamma_max):
try:
os.makedirs(out_folder)
except:
pass
test_fpath = os.path.join(folder, 'test.dat')
cents_fpath = os.path.join(folder, 'ksc', 'cents.dat')
assign_fpath = os.path.join(folder, 'ksc', 'test_assign.dat')
probs_folder = os.path.join(folder, 'probs-test')
X = ioutil.load_series(tseries_fpath, test_fpath)
test_idx = np.loadtxt(test_fpath, dtype='bool')
y_true = np.loadtxt(assign_fpath)
num_series = X.shape[0]
max_pts = gamma_max
peak_days = []
sum_views = []
with open(tseries_fpath) as tseries_file:
for i, line in enumerate(tseries_file):
if test_idx[i]:
x = np.array([int(v) for v in line.split()[1:]])
peak_days.append(x.argmax())
sum_views.append(x.sum())
peak_days = np.array(peak_days)
sum_views = np.array(sum_views)
y_pred, best_by, confs, all_confs = \
pred(probs_folder, num_series, max_pts, min_pts, thetas)
save_results(X, peak_days, sum_views, min_pts, thetas, best_by, all_confs,
y_true, y_pred, confs, out_folder)
def main(tseries_fpath, test_fpath, cents_fpath):
X = ioutil.load_series(tseries_fpath, test_fpath)
C = np.loadtxt(cents_fpath)
dist_cents = dist.dist_all(C, X, rolling=True)[0]
y_true = dist_cents.argmin(axis=0)
for t in y_true:
print t
def main(tseries_fpath, base_folder, k):
k = int(k)
idx_fpath = os.path.join(os.path.join(base_folder, ".."), "train.dat")
X = ioutil.load_series(tseries_fpath, idx_fpath)
cent, assign, shift, dists_cent = ksc.inc_ksc(X, k)
np.savetxt(os.path.join(base_folder, "cents.dat"), cent, fmt="%.5f")
np.savetxt(os.path.join(base_folder, "assign.dat"), assign, fmt="%d")
np.savetxt(os.path.join(base_folder, "shift.dat"), shift, fmt="%d")
np.savetxt(os.path.join(base_folder, "dists_cent.dat"), dists_cent, fmt="%.5f")
def main(tseries_fpath, base_folder, k):
k = int(k)
idx_fpath = os.path.join(os.path.join(base_folder, '..'), 'train.dat')
X = ioutil.load_series(tseries_fpath, idx_fpath)
cent, assign, shift, dists_cent = ksc.inc_ksc(X, k)
np.savetxt(os.path.join(base_folder, 'cents.dat'), cent, fmt='%.5f')
np.savetxt(os.path.join(base_folder, 'assign.dat'), assign, fmt='%d')
np.savetxt(os.path.join(base_folder, 'shift.dat'), shift, fmt='%d')
np.savetxt(os.path.join(base_folder, 'dists_cent.dat'), dists_cent,
fmt='%.5f')
def main(tseries_fpath, centroids_fpath, test_fpath, assign_fpath, out_folder):
C = np.genfromtxt(centroids_fpath)
Xtest = ioutil.load_series(tseries_fpath, test_fpath)
y_train = np.arange(C.shape[0])
max_pts = Xtest.shape[1]
for num_pts in range(1, max_pts + 1):
#for num_pts in [1, 25, 50, 75]:
probs = fit(C, y_train, Xtest, num_pts)
probs_fpath = os.path.join(out_folder, 'probs-%d-pts.dat' % num_pts)
np.savetxt(probs_fpath, probs)
def main(tseries_fpath, train_fpath, test_fpath, ytrain_fpath, ytest_fpath, out_folder):
Xtrain = ioutil.load_series(tseries_fpath, train_fpath)
Xtest = ioutil.load_series(tseries_fpath, test_fpath)
y_train = np.genfromtxt(ytrain_fpath)
y_true = np.genfromtxt(ytest_fpath)
max_pts = Xtrain.shape[1]
best_by = np.zeros(Xtest.shape[0])
min_conf = np.zeros(Xtest.shape[0])
all_probs = np.zeros(shape=(Xtest.shape[0], max_pts))
lousy_conf = 1.0 / len(set(y_train)) #if confidence is equal to this, classifier did nothing
for num_pts in range(1, max_pts + 1):
y_pred, probs = fit(Xtrain, y_train, Xtest, num_pts)
for i in range(Xtest.shape[0]):
p_true = probs[i, y_true[i]]
if best_by[i] == 0 and y_pred[i] == y_true[i] and p_true > lousy_conf:
best_by[i] = num_pts
min_conf[i] = probs[i, y_true[i]]
all_probs[i, num_pts - 1] = p_true
summary_fpath = os.path.join(out_folder,\
'class_summ-%d-pts.dat' % num_pts)
probs_fpath = os.path.join(out_folder, 'probs-%d-pts.dat' % num_pts)
with open(summary_fpath, 'w') as summary_file:
print(classification_report(y_true, y_pred), file=summary_file)
np.savetxt(probs_fpath, probs)
best_fpath = os.path.join(out_folder, 'best-by.dat')
conf_fpath = os.path.join(out_folder, 'conf.dat')
all_conf_fpath = os.path.join(out_folder, 'all-conf.dat')
np.savetxt(best_fpath, best_by)
np.savetxt(conf_fpath, min_conf)
np.savetxt(all_conf_fpath, np.asarray(all_probs))
def main(tseries_fpath, train_fpath, test_fpath, ytrain_fpath, ytest_fpath, out_folder):
Xtrain = ioutil.load_series(tseries_fpath, train_fpath)
Xtest = ioutil.load_series(tseries_fpath, test_fpath)
y_train = np.genfromtxt(ytrain_fpath)
y_true = np.genfromtxt(ytest_fpath)
max_pts = Xtrain.shape[1]
best_by = np.zeros(Xtest.shape[0])
min_conf = np.zeros(Xtest.shape[0])
all_probs = np.zeros(shape=(Xtest.shape[0], max_pts))
lousy_conf = 1.0 / len(set(y_train)) # if confidence is equal to this, classifier did nothing
for num_pts in range(1, max_pts + 1):
y_pred, probs = fit(Xtrain, y_train, Xtest, num_pts)
for i in xrange(Xtest.shape[0]):
p_true = probs[i, y_true[i]]
if best_by[i] == 0 and y_pred[i] == y_true[i] and p_true > lousy_conf:
best_by[i] = num_pts
min_conf[i] = probs[i, y_true[i]]
all_probs[i, num_pts - 1] = p_true
summary_fpath = os.path.join(out_folder, "class_summ-%d-pts.dat" % num_pts)
probs_fpath = os.path.join(out_folder, "probs-%d-pts.dat" % num_pts)
with open(summary_fpath, "w") as summary_file:
print(classification_report(y_true, y_pred), file=summary_file)
np.savetxt(probs_fpath, probs)
best_fpath = os.path.join(out_folder, "best-by.dat")
conf_fpath = os.path.join(out_folder, "conf.dat")
all_conf_fpath = os.path.join(out_folder, "all-conf.dat")
np.savetxt(best_fpath, best_by)
np.savetxt(conf_fpath, min_conf)
np.savetxt(all_conf_fpath, np.asarray(all_probs))
def main(tseries_fpath, centroids_fpath, test_fpath, assign_fpath, out_folder,
gamma_max):
gamma_max = int(gamma_max)
C = np.genfromtxt(centroids_fpath)
Xtest = ioutil.load_series(tseries_fpath, test_fpath)
y_train = np.arange(C.shape[0])
max_pts = gamma_max
for num_pts in range(1, max_pts + 1):
#for num_pts in [1, 25, 50, 75]:
probs = fit(C, y_train, Xtest, num_pts)
probs_fpath = os.path.join(out_folder, 'probs-%d-pts.dat' % num_pts)
np.savetxt(probs_fpath, probs)
def main(tseries_fpath, train_fpath, centroids_fpath, classes_fpath,
out_folder, gamma_max):
gamma_max = int(gamma_max)
X = ioutil.load_series(tseries_fpath, train_fpath)
C = np.genfromtxt(centroids_fpath, dtype='f')
y_train = np.arange(C.shape[0])
y_true = np.genfromtxt(classes_fpath, dtype='i')
max_pts = gamma_max
#max_pts = X.shape[1]
best_by = np.zeros(X.shape[0])
min_conf = np.zeros(X.shape[0])
all_probs = np.zeros(shape=(X.shape[0], max_pts))
lousy_conf = 1.0 / C.shape[
0] #if confidence is equal to this, classifier did nothing
for num_pts in range(1, max_pts + 1):
y_pred, probs = fit(C, y_train, X, y_true, num_pts)
for i in xrange(X.shape[0]):
p_true = probs[i, y_true[i]]
if best_by[i] == 0 and y_pred[i] == y_true[
i] and p_true > lousy_conf:
best_by[i] = num_pts
min_conf[i] = probs[i, y_true[i]]
all_probs[i, num_pts - 1] = p_true
summary_fpath = os.path.join(out_folder,\
'class_summ-%d-pts.dat' % num_pts)
probs_fpath = os.path.join(out_folder, 'probs-%d-pts.dat' % num_pts)
with open(summary_fpath, 'w') as summary_file:
print(classification_report(y_true, y_pred), file=summary_file)
np.savetxt(probs_fpath, probs)
best_fpath = os.path.join(out_folder, 'best-by.dat')
conf_fpath = os.path.join(out_folder, 'conf.dat')
all_conf_fpath = os.path.join(out_folder, 'all-conf.dat')
np.savetxt(best_fpath, best_by)
np.savetxt(conf_fpath, min_conf)
np.savetxt(all_conf_fpath, np.asarray(all_probs))
def main(tseries_fpath, train_fpath, centroids_fpath, classes_fpath, out_folder,
gamma_max):
gamma_max = int(gamma_max)
X = ioutil.load_series(tseries_fpath, train_fpath)
C = np.genfromtxt(centroids_fpath, dtype='f')
y_train = np.arange(C.shape[0])
y_true = np.genfromtxt(classes_fpath)
max_pts = gamma_max
#max_pts = X.shape[1]
best_by = np.zeros(X.shape[0])
min_conf = np.zeros(X.shape[0])
all_probs = np.zeros(shape=(X.shape[0], max_pts))
lousy_conf = 1.0 / C.shape[0] #if confidence is equal to this, classifier did nothing
for num_pts in range(1, max_pts + 1):
y_pred, probs = fit(C, y_train, X, y_true, num_pts)
for i in xrange(X.shape[0]):
p_true = probs[i, y_true[i]]
if best_by[i] == 0 and y_pred[i] == y_true[i] and p_true > lousy_conf:
best_by[i] = num_pts
min_conf[i] = probs[i, y_true[i]]
all_probs[i, num_pts - 1] = p_true
summary_fpath = os.path.join(out_folder,\
'class_summ-%d-pts.dat' % num_pts)
probs_fpath = os.path.join(out_folder, 'probs-%d-pts.dat' % num_pts)
with open(summary_fpath, 'w') as summary_file:
print(classification_report(y_true, y_pred), file=summary_file)
np.savetxt(probs_fpath, probs)
best_fpath = os.path.join(out_folder, 'best-by.dat')
conf_fpath = os.path.join(out_folder, 'conf.dat')
all_conf_fpath = os.path.join(out_folder, 'all-conf.dat')
np.savetxt(best_fpath, best_by)
np.savetxt(conf_fpath, min_conf)
np.savetxt(all_conf_fpath, np.asarray(all_probs))
