ENH: Allow adjust=False when times is provided (#59142)

* add adjust parameter to the ewma variable times test. Add tests for disallowed decay-specification parameters when times is specified and adjust=False * allow adjust=False when times is provided * re-calculate alpha each iteration for irregular-spaced time series * whatsnew entry for allowing adjust=False with times * pre-commit style fixes * reduce line lengths to comply with pre-commit * reduce line lengths and apply ruff-reformat changes

Author: tserrao

doc/source/whatsnew/v3.0.0.rst

@@ -42,6 +42,7 @@ Other enhancements
 - :func:`DataFrame.to_excel` argument ``merge_cells`` now accepts a value of ``"columns"`` to only merge :class:`MultiIndex` column header header cells (:issue:`35384`)
 - :meth:`DataFrame.corrwith` now accepts ``min_periods`` as optional arguments, as in :meth:`DataFrame.corr` and :meth:`Series.corr` (:issue:`9490`)
 - :meth:`DataFrame.cummin`, :meth:`DataFrame.cummax`, :meth:`DataFrame.cumprod` and :meth:`DataFrame.cumsum` methods now have a ``numeric_only`` parameter (:issue:`53072`)
+- :meth:`DataFrame.ewm` now allows ``adjust=False`` when ``times`` is provided (:issue:`54328`)
 - :meth:`DataFrame.fillna` and :meth:`Series.fillna` can now accept ``value=None``; for non-object dtype the corresponding NA value will be used (:issue:`57723`)
 - :meth:`DataFrame.pivot_table` and :func:`pivot_table` now allow the passing of keyword arguments to ``aggfunc`` through ``**kwargs`` (:issue:`57884`)
 - :meth:`Series.cummin` and :meth:`Series.cummax` now supports :class:`CategoricalDtype` (:issue:`52335`)

pandas/_libs/window/aggregations.pyx

@@ -1813,6 +1813,9 @@ def ewm(const float64_t[:] vals, const int64_t[:] start, const int64_t[:] end,
  if normalize:
  # avoid numerical errors on constant series
  if weighted != cur:
+ if not adjust and com == 1:
+ # update in case of irregular-interval series
+ new_wt = 1. - old_wt
  weighted = old_wt * weighted + new_wt * cur
  weighted /= (old_wt + new_wt)
  if adjust:

pandas/core/window/ewm.py

@@ -134,8 +134,10 @@ class ExponentialMovingWindow(BaseWindow):
  Provide exponentially weighted (EW) calculations.
 
  Exactly one of ``com``, ``span``, ``halflife``, or ``alpha`` must be
- provided if ``times`` is not provided. If ``times`` is provided,
+ provided if ``times`` is not provided. If ``times`` is provided and ``adjust=True``,
  ``halflife`` and one of ``com``, ``span`` or ``alpha`` may be provided.
+ If ``times`` is provided and ``adjust=False``, ``halflife`` must be the only
+ provided decay-specification parameter.
 
  Parameters
  ----------
@@ -358,8 +360,6 @@ def __init__(
  self.ignore_na = ignore_na
  self.times = times
  if self.times is not None:
- if not self.adjust:
- raise NotImplementedError("times is not supported with adjust=False.")
  times_dtype = getattr(self.times, "dtype", None)
  if not (
  is_datetime64_dtype(times_dtype)
@@ -376,6 +376,11 @@ def __init__(
  # Halflife is no longer applicable when calculating COM
  # But allow COM to still be calculated if the user passes other decay args
  if common.count_not_none(self.com, self.span, self.alpha) > 0:
+ if not self.adjust:
+ raise NotImplementedError(
+ "None of com, span, or alpha can be specified if "
+ "times is provided and adjust=False"
+ )
  self._com = get_center_of_mass(self.com, self.span, None, self.alpha)
  else:
  self._com = 1.0

pandas/core/window/numba_.py

@@ -149,6 +149,9 @@ def ewm(
  # note that len(deltas) = len(vals) - 1 and deltas[i]
  # is to be used in conjunction with vals[i+1]
  old_wt *= old_wt_factor ** deltas[start + j - 1]
+ if not adjust and com == 1:
+ # update in case of irregular-interval time series
+ new_wt = 1.0 - old_wt
  else:
  weighted = old_wt_factor * weighted
  if is_observation:
@@ -324,6 +327,9 @@ def ewm_table(
  # note that len(deltas) = len(vals) - 1 and deltas[i]
  # is to be used in conjunction with vals[i+1]
  old_wt[j] *= old_wt_factor ** deltas[i - 1]
+ if not adjust and com == 1:
+ # update in case of irregular-interval time series
+ new_wt = 1.0 - old_wt[j]
  else:
  weighted[j] = old_wt_factor * weighted[j]
  if is_observations[j]:

pandas/tests/window/test_ewm.py

@@ -102,7 +102,8 @@ def test_ewma_with_times_equal_spacing(halflife_with_times, times, min_periods):
  tm.assert_frame_equal(result, expected)
 
 
-def test_ewma_with_times_variable_spacing(tz_aware_fixture, unit):
+def test_ewma_with_times_variable_spacing(tz_aware_fixture, unit, adjust):
+ # GH 54328
  tz = tz_aware_fixture
  halflife = "23 days"
  times = (
@@ -112,8 +113,11 @@ def test_ewma_with_times_variable_spacing(tz_aware_fixture, unit):
  )
  data = np.arange(3)
  df = DataFrame(data)
- result = df.ewm(halflife=halflife, times=times).mean()
- expected = DataFrame([0.0, 0.5674161888241773, 1.545239952073459])
+ result = df.ewm(halflife=halflife, times=times, adjust=adjust).mean()
+ if adjust:
+ expected = DataFrame([0.0, 0.5674161888241773, 1.545239952073459])
+ else:
+ expected = DataFrame([0.0, 0.23762518642226227, 1.534926369128742])
  tm.assert_frame_equal(result, expected)
 
 
@@ -148,13 +152,56 @@ def test_ewm_getitem_attributes_retained(arg, adjust, ignore_na):
  assert result == expected
 
 
-def test_ewma_times_adjust_false_raises():
- # GH 40098
+def test_ewma_times_adjust_false_with_disallowed_com():
+ # GH 54328
+ with pytest.raises(
+ NotImplementedError,
+ match=(
+ "None of com, span, or alpha can be specified "
+ "if times is provided and adjust=False"
+ ),
+ ):
+ Series(range(1)).ewm(
+ 0.1,
+ adjust=False,
+ times=date_range("2000", freq="D", periods=1),
+ halflife="1D",
+ )
+
+
+def test_ewma_times_adjust_false_with_disallowed_alpha():
+ # GH 54328
  with pytest.raises(
- NotImplementedError, match="times is not supported with adjust=False."
+ NotImplementedError,
+ match=(
+ "None of com, span, or alpha can be specified "
+ "if times is provided and adjust=False"
+ ),
+ ):
+ Series(range(1)).ewm(
+ 0.1,
+ adjust=False,
+ times=date_range("2000", freq="D", periods=1),
+ alpha=0.5,
+ halflife="1D",
+ )
+
+
+def test_ewma_times_adjust_false_with_disallowed_span():
+ # GH 54328
+ with pytest.raises(
+ NotImplementedError,
+ match=(
+ "None of com, span, or alpha can be specified "
+ "if times is provided and adjust=False"
+ ),
  ):
  Series(range(1)).ewm(
- 0.1, adjust=False, times=date_range("2000", freq="D", periods=1)
+ 0.1,
+ adjust=False,
+ times=date_range("2000", freq="D", periods=1),
+ span=10,
+ halflife="1D",
  )