with timer("Loading data"):
if args.normalize_target:
test = load_data("test_nn_target_normalized_meter")
test_square_feet = load_data("test_clean")["square_feet"].values
else:
test = load_data("test_nn_meter")
test["target"] = -1
with timer("Predicting"):
test_preds = np.zeros(len(test))
for m in range(4):
print(m)
# get base file name
model_name = f"mlp-split_meter"
make_dir(f"{MODEL_PATH}/{model_name}")
# create sub model path
if args.normalize_target:
sub_model_path = f"{MODEL_PATH}/mlp-split_meter/target_normalization/meter_{m}"
else:
sub_model_path = f"{MODEL_PATH}/mlp-split_meter/no_normalization/meter_{m}"
# remove indices not in this meter
X = test.loc[test.meter == m, FEATURES + ["target"]]
print(f"split meter {m}: test size {len(X)}")
# load models
model_list = glob.glob(f"{sub_model_path}/*")
# predict
with timer("Preprocesing"):
for x in CAT_COLS:
train[x] = train[x].astype("category")
if args.normalize_target:
target_encode_cols = [x for x in train.columns if "gte" in x]
train[target_encode_cols] = train[target_encode_cols] / np.log1p(
train[["square_feet"]].values)
train["target"] = np.log1p(train["meter_reading"]) / np.log1p(
train["square_feet"])
else:
train["target"] = np.log1p(train["meter_reading"])
# get base file name
model_name = f"lgb-split_site"
make_dir(f"{MODEL_PATH}/{model_name}")
for seed in range(3):
for n_months in [1, 2, 3, 4, 5, 6]:
validation_months_list = get_validation_months(6)
for fold_, validation_months in enumerate(validation_months_list):
for s in range(16):
# create sub model path
if args.normalize_target:
sub_model_path = f"{MODEL_PATH}/{model_name}/target_normalization/site_{s}"
make_dir(sub_model_path)
else:
sub_model_path = f"{MODEL_PATH}/{model_name}/no_normalization/site_{s}"
make_dir(sub_model_path)