time_stock = 10
cycle = 0
print("{:~^{n}}".format(basename.capitalize(), n=50))
while cycle <= 1: # max_cycle:
begin = time.time()
vprint( verbose, "=========== " + basename.capitalize() +" Training cycle " + str(cycle) +" ================")
n_estimators = 10
if cycle:
n_estimators = int((time_budget / time_spent_last - 1) * n_estimators - time_stock)
# n_estimators = int(((time_budget / time_spent_last) - 1) * 5) # * 5 == aim to use 5/10 of the time budget
if n_estimators <= 0:
break
print("{} estimators".format(n_estimators))
K = D.info['target_num']
task = D.info['task']
autoML = OurAutoML(D.info).fit(D.data['X_train'], D.data['Y_train'], cv=2, n_estimators=n_estimators)
vprint( verbose, "[+] Fitting success, time spent so far %5.2f sec" % (time.time() - start))
# Make predictions
Y_valid = autoML.predict(D.data['X_valid'])
Y_test = autoML.predict(D.data['X_test'])
print("score: {} ({} s)".format(autoML.scores.mean(), "%5.2f"%(time.time() - start)))
vprint( verbose, "[+] Prediction success, time spent so far %5.2f sec" % (time.time() - start))
# Write results
filename_valid = basename + '_valid_' + str(cycle).zfill(3) + '.predict'
data_io.write(os.path.join(output_dir,filename_valid), Y_valid)
filename_test = basename + '_test_' + str(cycle).zfill(3) + '.predict'
data_io.write(os.path.join(output_dir,filename_test), Y_test)
vprint( verbose, "[+] Results saved, time spent so far %5.2f sec" % (time.time() - start))
# The model can also select its hyper-parameters based on other elements of info.
# vprint( verbose, "======== Creating model ==========")
# M = MyAutoML(D.info, verbose, debug_mode)
# print M
# ========= Iterating over learning cycles and keeping track of time
time_spent = time.time() - start
time_budget = time_budget - time_spent # Remove time spent so far
start = time.time() # Reset the counter
time_spent = 0 # Initialize time spent learning
time_spent_last = 0 # Initialize time spent learning
time_stock = 5
cycle = 0
print("{:~^{n}}".format(basename.capitalize(), n=50))
autoML = OurAutoML(D.info)
X_train, Y_train = D.data['X_train'], D.data['Y_train']
n_estimators = 500
autoML.preprocess_bin_cl(X_train, Y_train, n_estimators=int(n_estimators/2))
cycle=1
while cycle <= 1: # max_cycle:
begin = time.time()
vprint( verbose, "=========== " + basename.capitalize() +" Training cycle " + str(cycle) +" ================")
print("{} estimators".format(n_estimators))
prev_n_estimators = n_estimators
if _MODE == "TEST_SCORE":
autoML.fit_and_count_av_score(X_train, Y_train,
n_estimators=n_estimators,
test_size=0.4)
else:
