def test_add_time_features():
tran = transform.AddTimeFeatures(
start_field=FieldName.START,
target_field=FieldName.TARGET,
output_field="time_feat",
time_features=[
time_feature.DayOfWeek(),
time_feature.DayOfMonth(),
time_feature.MonthOfYear(),
],
pred_length=10,
)
tran2 = clone(
tran,
{
"time_features": [
time_feature.DayOfWeek(),
time_feature.DayOfMonth(),
]
},
)
assert equals(tran, clone(tran))
assert not equals(tran, tran2)
def test_chain():
chain = transform.Chain(trans=[
transform.AddTimeFeatures(
start_field=FieldName.START,
target_field=FieldName.TARGET,
output_field="time_feat",
time_features=[
time_feature.DayOfWeek(),
time_feature.DayOfMonth(),
time_feature.MonthOfYear(),
],
pred_length=10,
),
transform.AddAgeFeature(
target_field=FieldName.TARGET,
output_field="age",
pred_length=10,
log_scale=True,
),
transform.AddObservedValuesIndicator(target_field=FieldName.TARGET,
output_field="observed_values"),
])
assert equals(chain, clone(chain))
assert not equals(chain, clone(chain, {"trans": []}))
another_chain = transform.Chain(trans=[
transform.AddTimeFeatures(
start_field=FieldName.START,
target_field=FieldName.TARGET,
output_field="time_feat",
time_features=[
time_feature.DayOfWeek(),
time_feature.DayOfMonth(),
time_feature.MonthOfYear(),
],
pred_length=10,
),
transform.AddAgeFeature(
target_field=FieldName.TARGET,
output_field="age",
pred_length=10,
log_scale=False,
),
transform.AddObservedValuesIndicator(target_field=FieldName.TARGET,
output_field="observed_values"),
])
assert not equals(chain, another_chain)
def dsinfo(request):
from gluonts import time_feature
from gluonts.dataset.artificial import constant_dataset, default_synthetic
if request.param == "constant":
ds_info, train_ds, test_ds = constant_dataset()
return AttrDict(
name="constant",
cardinality=int(ds_info.metadata.feat_static_cat[0].cardinality),
freq=ds_info.metadata.freq,
num_parallel_samples=2,
prediction_length=ds_info.prediction_length,
# FIXME: Should time features should not be needed for GP
time_features=[time_feature.DayOfWeek(),
time_feature.HourOfDay()],
train_ds=train_ds,
test_ds=test_ds,
)
elif request.param == "synthetic":
ds_info, train_ds, test_ds = default_synthetic()
return AttrDict(
name="synthetic",
batch_size=32,
cardinality=int(ds_info.metadata.feat_static_cat[0].cardinality),
context_length=2,
freq=ds_info.metadata.freq,
prediction_length=ds_info.prediction_length,
num_parallel_samples=2,
train_ds=train_ds,
test_ds=test_ds,
time_features=None,
)
def test_Transformation():
train_length = 100
ds = gluonts.dataset.common.ListDataset(
[{"start": "2012-01-01", "target": [0.2] * train_length}], freq="1D"
)
pred_length = 10
t = transform.Chain(
trans=[
transform.AddTimeFeatures(
start_field=transform.FieldName.START,
target_field=transform.FieldName.TARGET,
output_field="time_feat",
time_features=[
time_feature.DayOfWeek(),
time_feature.DayOfMonth(),
time_feature.MonthOfYear(),
],
pred_length=pred_length,
),
transform.AddAgeFeature(
target_field=transform.FieldName.TARGET,
output_field="age",
pred_length=pred_length,
log_scale=True,
),
transform.AddObservedValuesIndicator(
target_field=transform.FieldName.TARGET,
output_field="observed_values",
),
transform.VstackFeatures(
output_field="dynamic_feat",
input_fields=["age", "time_feat"],
drop_inputs=True,
),
transform.InstanceSplitter(
target_field=transform.FieldName.TARGET,
is_pad_field=transform.FieldName.IS_PAD,
start_field=transform.FieldName.START,
forecast_start_field=transform.FieldName.FORECAST_START,
train_sampler=transform.ExpectedNumInstanceSampler(
num_instances=4
),
past_length=train_length,
future_length=pred_length,
time_series_fields=["dynamic_feat", "observed_values"],
),
]
)
assert_serializable(t)
for u in t(iter(ds), is_train=True):
print(u)
def test_AddTimeFeatures(start, target, is_train):
pred_length = 13
t = transform.AddTimeFeatures(
start_field=transform.FieldName.START,
target_field=transform.FieldName.TARGET,
output_field="myout",
pred_length=pred_length,
time_features=[time_feature.DayOfWeek(), time_feature.DayOfMonth()],
)
assert_serializable(t)
data = {"start": start, "target": target}
res = t.map_transform(data, is_train=is_train)
mat = res["myout"]
expected_length = len(target) + (0 if is_train else pred_length)
assert mat.shape == (2, expected_length)
tmp_idx = pd.date_range(
start=start, freq=start.freq, periods=expected_length
)
assert np.alltrue(mat[0] == time_feature.DayOfWeek()(tmp_idx))
assert np.alltrue(mat[1] == time_feature.DayOfMonth()(tmp_idx))
def test_add_method():
chain = transform.AddTimeFeatures(
start_field=FieldName.START,
target_field=FieldName.TARGET,
output_field="time_feat",
time_features=[
time_feature.DayOfWeek(),
time_feature.DayOfMonth(),
time_feature.MonthOfYear(),
],
pred_length=24,
) + transform.AddAgeFeature(
target_field=FieldName.TARGET,
output_field="age",
pred_length=24,
log_scale=True,
)
assert isinstance(chain, transform.Chain)
def test_multi_dim_transformation(is_train):
train_length = 10
first_dim = np.arange(1, 11, 1).tolist()
first_dim[-1] = "NaN"
second_dim = np.arange(11, 21, 1).tolist()
second_dim[0] = "NaN"
ds = gluonts.dataset.common.ListDataset(
data_iter=[{"start": "2012-01-01", "target": [first_dim, second_dim]}],
freq="1D",
one_dim_target=False,
)
pred_length = 2
# Looks weird - but this is necessary to assert the nan entries correctly.
first_dim[-1] = np.nan
second_dim[0] = np.nan
t = transform.Chain(
trans=[
transform.AddTimeFeatures(
start_field=transform.FieldName.START,
target_field=transform.FieldName.TARGET,
output_field="time_feat",
time_features=[
time_feature.DayOfWeek(),
time_feature.DayOfMonth(),
time_feature.MonthOfYear(),
],
pred_length=pred_length,
),
transform.AddAgeFeature(
target_field=transform.FieldName.TARGET,
output_field="age",
pred_length=pred_length,
log_scale=True,
),
transform.AddObservedValuesIndicator(
target_field=transform.FieldName.TARGET,
output_field="observed_values",
convert_nans=False,
),
transform.VstackFeatures(
output_field="dynamic_feat",
input_fields=["age", "time_feat"],
drop_inputs=True,
),
transform.InstanceSplitter(
target_field=transform.FieldName.TARGET,
is_pad_field=transform.FieldName.IS_PAD,
start_field=transform.FieldName.START,
forecast_start_field=transform.FieldName.FORECAST_START,
train_sampler=transform.ExpectedNumInstanceSampler(
num_instances=4
),
past_length=train_length,
future_length=pred_length,
time_series_fields=["dynamic_feat", "observed_values"],
output_NTC=False,
),
]
)
assert_serializable(t)
if is_train:
for u in t(iter(ds), is_train=True):
assert_shape(u["past_target"], (2, 10))
assert_shape(u["past_dynamic_feat"], (4, 10))
assert_shape(u["past_observed_values"], (2, 10))
assert_shape(u["future_target"], (2, 2))
assert_padded_array(
u["past_observed_values"],
np.array([[1.0] * 9 + [0.0], [0.0] + [1.0] * 9]),
u["past_is_pad"],
)
assert_padded_array(
u["past_target"],
np.array([first_dim, second_dim]),
u["past_is_pad"],
)
else:
for u in t(iter(ds), is_train=False):
assert_shape(u["past_target"], (2, 10))
assert_shape(u["past_dynamic_feat"], (4, 10))
assert_shape(u["past_observed_values"], (2, 10))
assert_shape(u["future_target"], (2, 0))
assert_padded_array(
u["past_observed_values"],
np.array([[1.0] * 9 + [0.0], [0.0] + [1.0] * 9]),
u["past_is_pad"],
)
assert_padded_array(
u["past_target"],
np.array([first_dim, second_dim]),
u["past_is_pad"],
)
from gluonts.model.gp_forecaster import GaussianProcessEstimator
from gluonts.model.predictor import Predictor
from gluonts.model.seasonal_naive import SeasonalNaiveEstimator
from gluonts.model.seq2seq import (
MQCNNEstimator,
MQRNNEstimator,
Seq2SeqEstimator,
)
from gluonts.model.simple_feedforward import SimpleFeedForwardEstimator
dataset_info, train_ds, test_ds = constant_dataset()
freq = dataset_info.metadata.time_granularity
prediction_length = dataset_info.prediction_length
cardinality = int(dataset_info.metadata.feat_static_cat[0].cardinality)
# FIXME: Should time features should not be needed for GP
time_features = [time_feature.DayOfWeek(), time_feature.HourOfDay()]
num_eval_samples = 2
epochs = 1
def seq2seq_base(seq2seq_model, hybridize: bool = True, batches_per_epoch=1):
return (
seq2seq_model,
dict(
ctx='cpu',
epochs=epochs,
learning_rate=1e-2,
hybridize=hybridize,
prediction_length=prediction_length,
context_length=prediction_length,
num_eval_samples=num_eval_samples,
