Skip to content

A programming framework for agentic AI

Change the upper bound for "lags" hyperparameter for sklearn forecast models (#437)

Kevin Chenc75f97b475ba
diff --git a/flaml/model.py b/flaml/model.py
index 0a1470656..c0ecfddcf 100644
--- a/flaml/model.py
+++ b/flaml/model.py
@@ -1813,7 +1813,10 @@ class TS_SKLearn(SKLearnEstimator):
"low_cost_init_value": False,
},
"lags": {
- "domain": tune.randint(lower=1, upper=data_size[0] - pred_horizon),
+ "domain": tune.randint(
+ lower=1, upper=int(np.sqrt(data_size[0]))
+
+ ),
"init_value": 3,
},
}
diff --git a/flaml/tune/sample.py b/flaml/tune/sample.py
index 113006fe1..c31f36622 100644
--- a/flaml/tune/sample.py
+++ b/flaml/tune/sample.py
@@ -38,6 +38,7 @@ logger = logging.getLogger(__name__)
try:
from ray.tune.sample import _BackwardsCompatibleNumpyRng
except ImportError:
+
class _BackwardsCompatibleNumpyRng:
"""Thin wrapper to ensure backwards compatibility between
new and old numpy randomness generators.
diff --git a/test/automl/test_forecast.py b/test/automl/test_forecast.py
index 3adfb1b45..1715969ef 100644
--- a/test/automl/test_forecast.py
+++ b/test/automl/test_forecast.py
@@ -121,6 +121,23 @@ def test_numpy():
print(automl.predict(12))
+def test_numpy_large():
+ import numpy as np
+ import pandas as pd
+ from flaml import AutoML
+
+ X_train = pd.date_range("2017-01-01", periods=70000, freq="T")
+ y_train = pd.DataFrame(np.random.randint(6500, 7500, 70000))
+ automl = AutoML()
+ automl.fit(
+ X_train=X_train[:-10].values, # a single column of timestamp
+ y_train=y_train[:-10].values, # value for each timestamp
+ period=10, # time horizon to forecast, e.g., 12 months
+ task="ts_forecast",
+ time_budget=10, # time budget in seconds
+ )
+
+
def load_multi_dataset():
"""multivariate time series forecasting dataset"""
import pandas as pd
@@ -429,4 +446,4 @@ if __name__ == "__main__":
test_multivariate_forecast_num(60)
test_multivariate_forecast_cat(60)
test_numpy()
- test_forecast_classification()
+ test_forecast_classification(60)