fix dtype for sine_timeseries (#2856)

Author: dennisbader

CHANGELOG.md

@@ -21,6 +21,7 @@ but cannot always guarantee backwards compatibility. Changes that may **break co
 
 - Fixed a bug in `SKLearnModel.get_estimator()` for univariate quantile models that use `multi_models=False` , where using `quantile` did not return the correct fitted quantile model / estimator. [#2838](https://github.com/unit8co/darts/pull/2838) by [Dennis Bader](https://github.com/dennisbader).
 - Fixed a bug in `LightGBMModel` and `CatBoostModel` when using component-specific lags and categorical features, where certain lag scenarios could result in incorrect categorical feature declaration. [#2852](https://github.com/unit8co/darts/pull/2852) by [Dennis Bader](https://github.com/dennisbader).
+- Fixed a bug in `darts.utils.timeseries_generation.sine_timeseries()`, where the returned series ignored the specified `dtype`. [#2856](https://github.com/unit8co/darts/pull/2856) by [Dennis Bader](https://github.com/dennisbader).
 - Removed `darts/tests` and `examples` from the Darts package distribution. These are only required for internal testing. [#2854](https://github.com/unit8co/darts/pull/2854) by [Dennis Bader](https://github.com/dennisbader).
 
 **Dependencies**

darts/tests/utils/test_timeseries_generation.py

@@ -1,3 +1,4 @@
+import itertools
 from typing import Union
 
 import numpy as np
@@ -697,3 +698,35 @@ def components_f(*args, **kwargs):
         series_new = series_new.with_static_covariates(static_covs)
         series_new = series_new.with_hierarchy({"new2": ["new1"]})
         assert series_new == series_renamed
+
+    @pytest.mark.parametrize(
+        "config",
+        itertools.product(
+            [np.float32, np.float64],
+            [
+                (autoregressive_timeseries, {"coef": [1.0]}, False),
+                (constant_timeseries, {}, False),
+                (datetime_attribute_timeseries, {"attribute": "dayofweek"}, True),
+                (gaussian_timeseries, {}, False),
+                (holidays_timeseries, {"country_code": "CH"}, True),
+                (linear_timeseries, {}, False),
+                (random_walk_timeseries, {}, False),
+                (sine_timeseries, {}, False),
+            ],
+        ),
+    )
+    def test_generation_dtype(self, config):
+        dtype, (gen_func, gen_kwargs, requires_idx) = config
+
+        if requires_idx:
+            gen_kwargs["time_index"] = generate_index(
+                start="2000-01-01", length=10, freq="D"
+            )
+        else:
+            gen_kwargs["length"] = 10
+
+        # generate a TimeSeries with the specified dtype
+        ts = gen_func(dtype=dtype, **gen_kwargs)
+
+        # check that the dtype of the values is as expected
+        assert ts.dtype == dtype

darts/utils/timeseries_generation.py

@@ -3,7 +3,6 @@
 --------------------------------
 """
 
-import math
 from collections.abc import Sequence
 from typing import Any, Callable, Optional, Union
 
@@ -210,12 +209,10 @@ def sine_timeseries(
         start=start, end=end, freq=freq, length=length, name=TIMES_NAME
     )
     values = np.array(range(len(index)), dtype=dtype)
-    f = np.vectorize(
-        lambda x: value_amplitude
-        * math.sin(2 * math.pi * value_frequency * x + value_phase)
+    values = (
+        value_amplitude * np.sin(2 * np.pi * value_frequency * values + value_phase)
         + value_y_offset
     )
-    values = f(values)
     return TimeSeries(
         times=index,
         values=values,
@@ -371,6 +368,7 @@ def autoregressive_timeseries(
     length: Optional[int] = None,
     freq: Union[str, int] = None,
     column_name: Optional[str] = "autoregressive",
+    dtype: np.dtype = np.float64,
 ) -> TimeSeries:
     """
     Creates a univariate, autoregressive TimeSeries whose values are calculated using specified coefficients `coef` and
@@ -402,6 +400,8 @@ def autoregressive_timeseries(
         The freq is optional for generating an integer index (if not specified, 1 is used).
     column_name
         Optionally, the name of the value column for the returned TimeSeries
+    dtype
+        The desired NumPy dtype (np.float32 or np.float64) for the resulting series
 
     Returns
     -------
@@ -411,7 +411,7 @@ def autoregressive_timeseries(
 
     # if no start values specified default to a list of 1s
     if start_values is None:
-        start_values = np.ones(len(coef))
+        start_values = np.ones(len(coef), dtype=dtype)
     else:
         raise_if_not(
             len(start_values) == len(coef),
@@ -422,7 +422,7 @@ def autoregressive_timeseries(
         start=start, end=end, freq=freq, length=length, name=TIMES_NAME
     )
 
-    values = np.empty(len(coef) + len(index))
+    values = np.empty(len(coef) + len(index), dtype=dtype)
     values[: len(coef)] = start_values
 
     for i in range(len(coef), len(coef) + len(index)):