REF: lazify relativedelta imports (pandas-dev#52659)

* REF: lazify relativedelta imports

* API: intentionally raise ValueError

* whatsnew

Author: jbrockmendel

doc/source/whatsnew/v2.1.0.rst

@@ -307,6 +307,8 @@ Datetimelike
 - Bug in :func:`date_range` when ``freq`` was a :class:`DateOffset` with ``nanoseconds`` (:issue:`46877`)
 - Bug in :meth:`Timestamp.round` with values close to the implementation bounds returning incorrect results instead of raising ``OutOfBoundsDatetime`` (:issue:`51494`)
 - Bug in :meth:`arrays.DatetimeArray.map` and :meth:`DatetimeIndex.map`, where the supplied callable operated array-wise instead of element-wise (:issue:`51977`)
+- Bug in parsing datetime strings with weekday but no day e.g. "2023 Sept Thu" incorrectly raising ``AttributeError`` instead of ``ValueError`` (:issue:`52659`)
+-
 
 Timedelta
 ^^^^^^^^^

pandas/_libs/tslibs/offsets.pyx

@@ -15,8 +15,6 @@ from cpython.datetime cimport (
 
 import_datetime()
 
-from dateutil.easter import easter
-from dateutil.relativedelta import relativedelta
 import numpy as np
 
 cimport numpy as cnp
@@ -348,6 +346,8 @@ cdef _determine_offset(kwds):
         kwds_no_nanos["microseconds"] = kwds_no_nanos.get("microseconds", 0) + micro
 
     if all(k in kwds_use_relativedelta for k in kwds_no_nanos):
+        from dateutil.relativedelta import relativedelta
+
         return relativedelta(**kwds_no_nanos), True
 
     raise ValueError(
@@ -3691,6 +3691,8 @@ cdef class Easter(SingleConstructorOffset):
 
     @apply_wraps
     def _apply(self, other: datetime) -> datetime:
+        from dateutil.easter import easter
+
         current_easter = easter(other.year)
         current_easter = datetime(
             current_easter.year, current_easter.month, current_easter.day
@@ -3721,6 +3723,9 @@ cdef class Easter(SingleConstructorOffset):
     def is_on_offset(self, dt: datetime) -> bool:
         if self.normalize and not _is_normalized(dt):
             return False
+
+        from dateutil.easter import easter
+
         return date(dt.year, dt.month, dt.day) == easter(dt.year)
 
 

pandas/_libs/tslibs/parsing.pyx

@@ -46,7 +46,6 @@ from dateutil.parser import (
     DEFAULTPARSER,
     parse as du_parse,
 )
-from dateutil.relativedelta import relativedelta
 from dateutil.tz import (
     tzlocal as _dateutil_tzlocal,
     tzoffset,
@@ -692,7 +691,11 @@ cdef datetime dateutil_parse(
         ) from err
 
     if res.weekday is not None and not res.day:
-        ret = ret + relativedelta.relativedelta(weekday=res.weekday)
+        # GH#52659
+        raise ValueError(
+            "Parsing datetimes with weekday but no day information is "
+            "not supported"
+        )
     if not ignoretz:
         if res.tzname and res.tzname in time.tzname:
             # GH#50791

pandas/io/stata.py

@@ -12,7 +12,10 @@
 from __future__ import annotations
 
 from collections import abc
-import datetime
+from datetime import (
+    datetime,
+    timedelta,
+)
 from io import BytesIO
 import os
 import struct
@@ -30,7 +33,6 @@
 )
 import warnings
 
-from dateutil.relativedelta import relativedelta
 import numpy as np
 
 from pandas._libs import lib
@@ -226,7 +228,7 @@
 _date_formats = ["%tc", "%tC", "%td", "%d", "%tw", "%tm", "%tq", "%th", "%ty"]
 
 
-stata_epoch: Final = datetime.datetime(1960, 1, 1)
+stata_epoch: Final = datetime(1960, 1, 1)
 
 
 # TODO: Add typing. As of January 2020 it is not possible to type this function since
@@ -279,8 +281,8 @@ def _stata_elapsed_date_to_datetime_vec(dates, fmt) -> Series:
         years since 0000
     """
     MIN_YEAR, MAX_YEAR = Timestamp.min.year, Timestamp.max.year
-    MAX_DAY_DELTA = (Timestamp.max - datetime.datetime(1960, 1, 1)).days
-    MIN_DAY_DELTA = (Timestamp.min - datetime.datetime(1960, 1, 1)).days
+    MAX_DAY_DELTA = (Timestamp.max - datetime(1960, 1, 1)).days
+    MIN_DAY_DELTA = (Timestamp.min - datetime(1960, 1, 1)).days
     MIN_MS_DELTA = MIN_DAY_DELTA * 24 * 3600 * 1000
     MAX_MS_DELTA = MAX_DAY_DELTA * 24 * 3600 * 1000
 
@@ -295,9 +297,7 @@ def convert_year_month_safe(year, month) -> Series:
             return to_datetime(100 * year + month, format="%Y%m")
         else:
             index = getattr(year, "index", None)
-            return Series(
-                [datetime.datetime(y, m, 1) for y, m in zip(year, month)], index=index
-            )
+            return Series([datetime(y, m, 1) for y, m in zip(year, month)], index=index)
 
     def convert_year_days_safe(year, days) -> Series:
         """
@@ -309,8 +309,7 @@ def convert_year_days_safe(year, days) -> Series:
         else:
             index = getattr(year, "index", None)
             value = [
-                datetime.datetime(y, 1, 1) + relativedelta(days=int(d))
-                for y, d in zip(year, days)
+                datetime(y, 1, 1) + timedelta(days=int(d)) for y, d in zip(year, days)
             ]
             return Series(value, index=index)
 
@@ -323,12 +322,12 @@ def convert_delta_safe(base, deltas, unit) -> Series:
         index = getattr(deltas, "index", None)
         if unit == "d":
             if deltas.max() > MAX_DAY_DELTA or deltas.min() < MIN_DAY_DELTA:
-                values = [base + relativedelta(days=int(d)) for d in deltas]
+                values = [base + timedelta(days=int(d)) for d in deltas]
                 return Series(values, index=index)
         elif unit == "ms":
             if deltas.max() > MAX_MS_DELTA or deltas.min() < MIN_MS_DELTA:
                 values = [
-                    base + relativedelta(microseconds=(int(d) * 1000)) for d in deltas
+                    base + timedelta(microseconds=(int(d) * 1000)) for d in deltas
                 ]
                 return Series(values, index=index)
         else:
@@ -405,7 +404,7 @@ def _datetime_to_stata_elapsed_vec(dates: Series, fmt: str) -> Series:
     Parameters
     ----------
     dates : Series
-        Series or array containing datetime.datetime or datetime64[ns] to
+        Series or array containing datetime or datetime64[ns] to
         convert to the Stata Internal Format given by fmt
     fmt : str
         The format to convert to. Can be, tc, td, tw, tm, tq, th, ty
@@ -436,7 +435,7 @@ def parse_dates_safe(
             if delta:
                 delta = dates._values - stata_epoch
 
-                def f(x: datetime.timedelta) -> float:
+                def f(x: timedelta) -> float:
                     return US_PER_DAY * x.days + 1000000 * x.seconds + x.microseconds
 
                 v = np.vectorize(f)
@@ -447,15 +446,15 @@ def f(x: datetime.timedelta) -> float:
                 d["month"] = year_month._values - d["year"] * 100
             if days:
 
-                def g(x: datetime.datetime) -> int:
-                    return (x - datetime.datetime(x.year, 1, 1)).days
+                def g(x: datetime) -> int:
+                    return (x - datetime(x.year, 1, 1)).days
 
                 v = np.vectorize(g)
                 d["days"] = v(dates)
         else:
             raise ValueError(
                 "Columns containing dates must contain either "
-                "datetime64, datetime.datetime or null values."
+                "datetime64, datetime or null values."
             )
 
         return DataFrame(d, index=index)
@@ -2291,7 +2290,7 @@ class StataWriter(StataParser):
         * If datetimes contain timezone information
     ValueError
         * Columns listed in convert_dates are neither datetime64[ns]
-          or datetime.datetime
+          or datetime
         * Column dtype is not representable in Stata
         * Column listed in convert_dates is not in DataFrame
         * Categorical label contains more than 32,000 characters
@@ -2324,7 +2323,7 @@ def __init__(
         convert_dates: dict[Hashable, str] | None = None,
         write_index: bool = True,
         byteorder: str | None = None,
-        time_stamp: datetime.datetime | None = None,
+        time_stamp: datetime | None = None,
         data_label: str | None = None,
         variable_labels: dict[Hashable, str] | None = None,
         compression: CompressionOptions = "infer",
@@ -2764,7 +2763,7 @@ def _write_value_labels(self) -> None:
     def _write_header(
         self,
         data_label: str | None = None,
-        time_stamp: datetime.datetime | None = None,
+        time_stamp: datetime | None = None,
     ) -> None:
         byteorder = self._byteorder
         # ds_format - just use 114
@@ -2789,8 +2788,8 @@ def _write_header(
         # time stamp, 18 bytes, char, null terminated
         # format dd Mon yyyy hh:mm
         if time_stamp is None:
-            time_stamp = datetime.datetime.now()
-        elif not isinstance(time_stamp, datetime.datetime):
+            time_stamp = datetime.now()
+        elif not isinstance(time_stamp, datetime):
             raise ValueError("time_stamp should be datetime type")
         # GH #13856
         # Avoid locale-specific month conversion
@@ -3214,7 +3213,7 @@ class StataWriter117(StataWriter):
         * If datetimes contain timezone information
     ValueError
         * Columns listed in convert_dates are neither datetime64[ns]
-          or datetime.datetime
+          or datetime
         * Column dtype is not representable in Stata
         * Column listed in convert_dates is not in DataFrame
         * Categorical label contains more than 32,000 characters
@@ -3250,7 +3249,7 @@ def __init__(
         convert_dates: dict[Hashable, str] | None = None,
         write_index: bool = True,
         byteorder: str | None = None,
-        time_stamp: datetime.datetime | None = None,
+        time_stamp: datetime | None = None,
         data_label: str | None = None,
         variable_labels: dict[Hashable, str] | None = None,
         convert_strl: Sequence[Hashable] | None = None,
@@ -3295,7 +3294,7 @@ def _update_map(self, tag: str) -> None:
     def _write_header(
         self,
         data_label: str | None = None,
-        time_stamp: datetime.datetime | None = None,
+        time_stamp: datetime | None = None,
     ) -> None:
         """Write the file header"""
         byteorder = self._byteorder
@@ -3321,8 +3320,8 @@ def _write_header(
         # time stamp, 18 bytes, char, null terminated
         # format dd Mon yyyy hh:mm
         if time_stamp is None:
-            time_stamp = datetime.datetime.now()
-        elif not isinstance(time_stamp, datetime.datetime):
+            time_stamp = datetime.now()
+        elif not isinstance(time_stamp, datetime):
             raise ValueError("time_stamp should be datetime type")
         # Avoid locale-specific month conversion
         months = [
@@ -3604,7 +3603,7 @@ class StataWriterUTF8(StataWriter117):
         * If datetimes contain timezone information
     ValueError
         * Columns listed in convert_dates are neither datetime64[ns]
-          or datetime.datetime
+          or datetime
         * Column dtype is not representable in Stata
         * Column listed in convert_dates is not in DataFrame
         * Categorical label contains more than 32,000 characters
@@ -3641,7 +3640,7 @@ def __init__(
         convert_dates: dict[Hashable, str] | None = None,
         write_index: bool = True,
         byteorder: str | None = None,
-        time_stamp: datetime.datetime | None = None,
+        time_stamp: datetime | None = None,
         data_label: str | None = None,
         variable_labels: dict[Hashable, str] | None = None,
         convert_strl: Sequence[Hashable] | None = None,

pandas/plotting/_matplotlib/converter.py

@@ -16,7 +16,6 @@
     cast,
 )
 
-from dateutil.relativedelta import relativedelta
 import matplotlib.dates as mdates
 from matplotlib.ticker import (
     AutoLocator,
@@ -349,11 +348,7 @@ def __init__(self, locator, tz=None, defaultfmt: str = "%Y-%m-%d") -> None:
 class PandasAutoDateLocator(mdates.AutoDateLocator):
     def get_locator(self, dmin, dmax):
         """Pick the best locator based on a distance."""
-        delta = relativedelta(dmax, dmin)
-
-        num_days = (delta.years * 12.0 + delta.months) * 31.0 + delta.days
-        num_sec = (delta.hours * 60.0 + delta.minutes) * 60.0 + delta.seconds
-        tot_sec = num_days * 86400.0 + num_sec
+        tot_sec = (dmax - dmin).total_seconds()
 
         if abs(tot_sec) < self.minticks:
             self._freq = -1

pandas/tests/scalar/timestamp/test_constructors.py

@@ -25,6 +25,12 @@
 
 
 class TestTimestampConstructors:
+    def test_weekday_but_no_day_raises(self):
+        # GH#52659
+        msg = "Parsing datetimes with weekday but no day information is not supported"
+        with pytest.raises(ValueError, match=msg):
+            Timestamp("2023 Sept Thu")
+
     def test_construct_from_string_invalid_raises(self):
         # dateutil (weirdly) parses "200622-12-31" as
         #  datetime(2022, 6, 20, 12, 0, tzinfo=tzoffset(None, -111600)