pandas-market-calendars 4.3.3__py3-none-any.whl → 4.6.0__py3-none-any.whl

pandas_market_calendars/market_calendar.py

@@ -1,895 +1,1057 @@
-# Fork of zipline from Quantopian. Licensed under MIT, original licence below
-#
-# Copyright 2016 Quantopian, Inc.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-import warnings
-from abc import ABCMeta, abstractmethod
-from datetime import time
-
-import pandas as pd
-from pandas.tseries.offsets import CustomBusinessDay
-
-from .class_registry import RegisteryMeta, ProtectedDict
-
-MONDAY, TUESDAY, WEDNESDAY, THURSDAY, FRIDAY, SATURDAY, SUNDAY = range(7)
-
-
-class DEFAULT:
-    pass
-
-
-class MarketCalendarMeta(ABCMeta, RegisteryMeta):
-    pass
-
-
-class MarketCalendar(metaclass=MarketCalendarMeta):
-    """
-    An MarketCalendar represents the timing information of a single market or exchange.
-    Unless otherwise noted all times are in UTC and use Pandas data structures.
-    """
-
-    regular_market_times = {
-        "market_open": ((None, time(0)),),
-        "market_close": ((None, time(23)),),
-    }
-
-    open_close_map = {
-        "market_open": True,
-        "market_close": False,
-        "break_start": False,
-        "break_end": True,
-        "pre": True,
-        "post": False,
-    }
-
-    @staticmethod
-    def _tdelta(t, day_offset=0):
-        try:
-            return pd.Timedelta(
-                days=day_offset, hours=t.hour, minutes=t.minute, seconds=t.second
-            )
-        except AttributeError:
-            t, day_offset = t
-            return pd.Timedelta(
-                days=day_offset, hours=t.hour, minutes=t.minute, seconds=t.second
-            )
-
-    @staticmethod
-    def _off(tple):
-        try:
-            return tple[2]
-        except IndexError:
-            return 0
-
-    @classmethod
-    def calendar_names(cls):
-        """All Market Calendar names and aliases that can be used in "factory"
-        :return: list(str)
-        """
-        return [
-            cal
-            for cal in cls._regmeta_class_registry.keys()
-            if cal not in ["MarketCalendar", "TradingCalendar"]
-        ]
-
-    @classmethod
-    def factory(
-        cls, name, *args, **kwargs
-    ):  # Will be set by Meta, keeping it there for tests
-        """
-        :param name: The name of the MarketCalendar to be retrieved.
-        :param *args/**kwargs: passed to requested MarketCalendar.__init__
-        :return: MarketCalendar of the desired calendar.
-        """
-        return
-
-    def __init__(self, open_time=None, close_time=None):
-        """
-        :param open_time: Market open time override as datetime.time object. If None then default is used.
-        :param close_time: Market close time override as datetime.time object. If None then default is used.
-        """
-
-        self.regular_market_times = self.regular_market_times.copy()
-        self.open_close_map = self.open_close_map.copy()
-        self._customized_market_times = []
-
-        if open_time is not None:
-            self.change_time("market_open", open_time)
-
-        if close_time is not None:
-            self.change_time("market_close", close_time)
-
-        if not hasattr(self, "_market_times"):
-            self._prepare_regular_market_times()
-
-    @property
-    @abstractmethod
-    def name(self):
-        """
-        Name of the market
-
-        :return: string name
-        """
-        raise NotImplementedError()
-
-    @property
-    @abstractmethod
-    def tz(self):
-        """
-        Time zone for the market.
-
-        :return: timezone
-        """
-        raise NotImplementedError()
-
-    @property
-    def market_times(self):
-        return self._market_times
-
-    def _prepare_regular_market_times(self):
-        oc_map = self.open_close_map
-        assert all(
-            isinstance(x, bool) for x in oc_map.values()
-        ), "Values in open_close_map need to be True or False"
-
-        regular = self.regular_market_times
-        discontinued = ProtectedDict()
-        regular_tds = {}
-
-        for market_time, times in regular.items():
-            # in case a market_time has been discontinued, extend the last time
-            # and add it to the discontinued_market_times dictionary
-            if market_time.startswith("interruption_"):
-                raise ValueError("'interruption_' prefix is reserved")
-
-            if times[-1][1] is None:
-                discontinued._set(market_time, times[-1][0])
-                times = times[:-1]
-                regular._set(market_time, times)
-
-            regular_tds[market_time] = tuple(
-                (t[0], self._tdelta(t[1], self._off(t))) for t in times
-            )
-
-        if discontinued:
-            warnings.warn(
-                f"{list(discontinued.keys())} are discontinued, the dictionary"
-                f" `.discontinued_market_times` has the dates on which these were discontinued."
-                f" The times as of those dates are incorrect, use .remove_time(market_time)"
-                f" to ignore a market_time."
-            )
-
-        self.discontinued_market_times = discontinued
-        self.regular_market_times = regular
-
-        self._regular_market_timedeltas = regular_tds
-        self._market_times = sorted(regular.keys(), key=lambda x: regular_tds[x][-1][1])
-        self._oc_market_times = list(filter(oc_map.__contains__, self._market_times))
-
-    def _set_time(self, market_time, times, opens):
-        if isinstance(times, (tuple, list)):  # passed a tuple
-            if not isinstance(times[0], (tuple, list)):  # doesn't have a tuple inside
-                if times[0] is None:  # seems to be a tuple indicating starting time
-                    times = (times,)
-                else:  # must be a tuple with: (time, offset)
-                    times = ((None, times[0], times[1]),)
-        else:  # should be a datetime.time object
-            times = ((None, times),)
-
-        ln = len(times)
-        for i, t in enumerate(times):
-            try:
-                assert (
-                    t[0] is None
-                    or isinstance(t[0], str)
-                    or isinstance(t[0], pd.Timestamp)
-                )
-                assert isinstance(t[1], time) or (
-                    ln > 1 and i == ln - 1 and t[1] is None
-                )
-                assert isinstance(self._off(t), int)
-            except AssertionError:
-                raise AssertionError(
-                    "The passed time information is not in the right format, "
-                    "please consult the docs for how to set market times"
-                )
-
-        if opens is DEFAULT:
-            opens = self.__class__.open_close_map.get(market_time, None)
-
-        if opens in (True, False):
-            self.open_close_map._set(market_time, opens)
-
-        elif opens is None:  # make sure it's ignored
-            try:
-                self.open_close_map._del(market_time)
-            except KeyError:
-                pass
-        else:
-            raise ValueError(
-                "when you pass `opens`, it needs to be True, False, or None"
-            )
-
-        self.regular_market_times._set(market_time, times)
-
-        if not self.is_custom(market_time):
-            self._customized_market_times.append(market_time)
-
-        self._prepare_regular_market_times()
-
-    def change_time(self, market_time, times, opens=DEFAULT):
-        """
-        Changes the specified market time in regular_market_times and makes the necessary adjustments.
-
-        :param market_time: the market_time to change
-        :param times: new time information
-        :param opens: whether the market_time is a time that closes or opens the market
-            this is only needed if the market_time should be respected by .open_at_time
-            True: opens
-            False: closes
-            None: consider it neither opening nor closing, don't add to open_close_map (ignore in .open_at_time)
-            DEFAULT: same as None, unless the market_time is in self.__class__.open_close_map. Then it will take
-                the default value as defined by the class.
-        :return: None
-        """
-        assert market_time in self.regular_market_times, (
-            f"{market_time} is not in regular_market_times:" f"\n{self._market_times}."
-        )
-        return self._set_time(market_time, times, opens)
-
-    def add_time(self, market_time, times, opens=DEFAULT):
-        """
-        Adds the specified market time to regular_market_times and makes the necessary adjustments.
-
-        :param market_time: the market_time to add
-        :param times: the time information
-        :param opens: see .change_time docstring
-        :return: None
-        """
-        assert market_time not in self.regular_market_times, (
-            f"{market_time} is already in regular_market_times:"
-            f"\n{self._market_times}"
-        )
-
-        return self._set_time(market_time, times, opens)
-
-    def remove_time(self, market_time):
-        """
-        Removes the specified market time from regular_market_times and makes the necessary adjustments.
-
-        :param market_time: the market_time to remove
-        :return: None
-        """
-
-        self.regular_market_times._del(market_time)
-        try:
-            self.open_close_map._del(market_time)
-        except KeyError:
-            pass
-
-        self._prepare_regular_market_times()
-        if self.is_custom(market_time):
-            self._customized_market_times.remove(market_time)
-
-    def is_custom(self, market_time):
-        return market_time in self._customized_market_times
-
-    @property
-    def has_custom(self):
-        return len(self._customized_market_times) > 0
-
-    def is_discontinued(self, market_time):
-        return market_time in self.discontinued_market_times
-
-    @property
-    def has_discontinued(self):
-        return len(self.discontinued_market_times) > 0
-
-    def get_time(self, market_time, all_times=False):
-        try:
-            times = self.regular_market_times[market_time]
-        except KeyError as e:
-            if "break_start" in market_time or "break_end" in market_time:
-                return None  # in case of no breaks
-            elif market_time in ["market_open", "market_close"]:
-                raise NotImplementedError("You need to set market_times")
-            else:
-                raise e
-
-        if all_times:
-            return times
-        return times[-1][1].replace(tzinfo=self.tz)
-
-    def get_time_on(self, market_time, date):
-        times = self.get_time(market_time, all_times=True)
-        if times is None:
-            return None
-
-        date = pd.Timestamp(date)
-        for d, t in times[::-1]:
-            if d is None or pd.Timestamp(d) < date:
-                return t.replace(tzinfo=self.tz)
-
-    def open_time_on(self, date):
-        return self.get_time_on("market_open", date)
-
-    def close_time_on(self, date):
-        return self.get_time_on("market_close", date)
-
-    def break_start_on(self, date):
-        return self.get_time_on("break_start", date)
-
-    def break_end_on(self, date):
-        return self.get_time_on("break_end", date)
-
-    @property
-    def open_time(self):
-        """
-        Default open time for the market
-
-        :return: time
-        """
-        return self.get_time("market_open")
-
-    @property
-    def close_time(self):
-        """
-        Default close time for the market
-
-        :return: time
-        """
-        return self.get_time("market_close")
-
-    @property
-    def break_start(self):
-        """
-        Break time start. If None then there is no break
-
-        :return: time or None
-        """
-        return self.get_time("break_start")
-
-    @property
-    def break_end(self):
-        """
-        Break time end. If None then there is no break
-
-        :return: time or None
-        """
-        return self.get_time("break_end")
-
-    @property
-    def regular_holidays(self):
-        """
-
-        :return: pd.AbstractHolidayCalendar: a calendar containing the regular holidays for this calendar
-        """
-        return None
-
-    @property
-    def adhoc_holidays(self):
-        """
-
-        :return: list of ad-hoc holidays
-        """
-        return []
-
-    @property
-    def weekmask(self):
-        return "Mon Tue Wed Thu Fri"
-
-    @property
-    def special_opens(self):
-        """
-        A list of special open times and corresponding AbstractHolidayCalendar.
-
-        :return: List of (time, AbstractHolidayCalendar) tuples
-        """
-        return []
-
-    @property
-    def special_opens_adhoc(self):
-        """
-
-        :return: List of (time, DatetimeIndex) tuples that represent special opens that cannot be codified into rules.
-        """
-        return []
-
-    @property
-    def special_closes(self):
-        """
-        A list of special close times and corresponding HolidayCalendars.
-
-        :return: List of (time, AbstractHolidayCalendar) tuples
-        """
-        return []
-
-    @property
-    def special_closes_adhoc(self):
-        """
-
-        :return: List of (time, DatetimeIndex) tuples that represent special closes that cannot be codified into rules.
-        """
-        return []
-
-    def get_special_times(self, market_time):
-        return getattr(self, "special_" + market_time, [])
-
-    def get_special_times_adhoc(self, market_time):
-        return getattr(self, "special_" + market_time + "_adhoc", [])
-
-    def get_offset(self, market_time):
-        return self._off(self.get_time(market_time, all_times=True)[-1])
-
-    @property
-    def open_offset(self):
-        """
-        :return: open offset
-        """
-        return self.get_offset("market_open")
-
-    @property
-    def close_offset(self):
-        """
-        :return: close offset
-        """
-        return self.get_offset("market_close")
-
-    @property
-    def interruptions(self):
-        """
-        This needs to be a list with a tuple for each date that had an interruption.
-        The tuple should have this layout:
-
-            (date, start_time, end_time[, start_time2, end_time2, ...])
-
-        E.g.:
-        [
-            ("2002-02-03", (time(11), -1), time(11, 2)),
-            ("2010-01-11", time(11), (time(11, 1), 1)),
-            ("2010-01-13", time(9, 59), time(10), time(10, 29), time(10, 30)),
-            ("2011-01-10", time(11), time(11, 1))
-        ]
-
-        The date needs to be a string in this format: 'yyyy-mm-dd'.
-        Times need to be two datetime.time objects for each interruption, indicating start and end.
-         Optionally these can be wrapped in a tuple, where the
-         second element needs to be an integer indicating an offset.
-        On "2010-01-13" in the example, it is shown that there can be multiple interruptions in a day.
-        """
-        return []
-
-    def _convert(self, col: pd.Series):
-        """
-        col is a series indexed by dates at which interruptions occurred. The values are either the start or end times
-        of an interruption, represented by either a timedelta or a tuple with a timedelta and day offset of the form
-        (timedelta, offset). _convert produces a new series where the values are replaced by datetimes equal to the
-        index of the original series plus the offset if present, at the timedelta.
-
-        E.g.:
-        >>> self._convert(
-                pd.Series(
-                    [datetime.time(11, 2), (datetime.time(11, 1), 1), datetime.time(10, 0), None],
-                    index=pd.DatetimeIndex(['2002-02-03', '2010-01-11', '2010-01-13', '2011-01-10'])
-                )
-            )
-        2002-02-03   2002-02-03 11:02:00+00:00
-        2010-01-11   2010-01-12 11:01:00+00:00
-        2010-01-13   2010-01-13 10:00:00+00:00
-        2011-01-10                         NaT
-        dtype: datetime64[ns, UTC]
-        """
-        col = (
-            col.dropna()
-        )  # Python 3.8, pandas 2.0.3 cannot create time deltas from NaT
-        try:
-            times = col.str[0]
-        except AttributeError:  # no tuples, only offset 0
-            return (
-                (
-                    pd.to_timedelta(col.astype("string").fillna(""), errors="coerce")
-                    + col.index
-                )
-                .dt.tz_localize(self.tz)
-                .dt.tz_convert("UTC")
-            )
-
-        return (
-            (
-                pd.to_timedelta(
-                    times.fillna(col).astype("string").fillna(""), errors="coerce"
-                )
-                + pd.to_timedelta(col.str[1].fillna(0), unit="D")
-                + col.index
-            )
-            .dt.tz_localize(self.tz)
-            .dt.tz_convert("UTC")
-        )
-
-    @staticmethod
-    def _col_name(n: int):
-        return (
-            f"interruption_start_{n // 2 + 1}"
-            if n % 2 == 1
-            else f"interruption_end_{n // 2}"
-        )
-
-    @property
-    def interruptions_df(self):
-        """
-        Will return a pd.DataFrame only containing interruptions.
-        """
-        if not self.interruptions:
-            return pd.DataFrame(index=pd.DatetimeIndex([]))
-        intr = pd.DataFrame(self.interruptions)
-        intr.index = pd.to_datetime(intr.pop(0))
-
-        intr.columns = map(self._col_name, intr.columns)
-        intr.index.name = None
-
-        return intr.apply(self._convert).sort_index()
-
-    def holidays(self):
-        """
-        Returns the complete CustomBusinessDay object of holidays that can be used in any Pandas function that take
-        that input.
-
-        :return: CustomBusinessDay object of holidays
-        """
-        try:
-            return self._holidays
-        except AttributeError:
-            self._holidays = CustomBusinessDay(
-                holidays=self.adhoc_holidays,
-                calendar=self.regular_holidays,
-                weekmask=self.weekmask,
-            )
-        return self._holidays
-
-    def valid_days(self, start_date, end_date, tz="UTC"):
-        """
-        Get a DatetimeIndex of valid open business days.
-
-        :param start_date: start date
-        :param end_date: end date
-        :param tz: time zone in either string or pytz.timezone
-        :return: DatetimeIndex of valid business days
-        """
-        return pd.date_range(
-            start_date, end_date, freq=self.holidays(), normalize=True, tz=tz
-        )
-
-    def _get_market_times(self, start, end):
-        mts = self._market_times
-        return mts[mts.index(start) : mts.index(end) + 1]
-
-    def days_at_time(self, days, market_time, day_offset=0):
-        """
-        Create an index of days at time ``t``, interpreted in timezone ``tz``. The returned index is localized to UTC.
-
-        In the example below, the times switch from 13:45 to 12:45 UTC because
-        March 13th is the daylight savings transition for US/Eastern.  All the
-        times are still 8:45 when interpreted in US/Eastern.
-
-        >>> import pandas as pd; import datetime; import pprint
-        >>> dts = pd.date_range('2016-03-12', '2016-03-14')
-        >>> dts_at_845 = days_at_time(dts, datetime.time(8, 45), 'US/Eastern')
-        >>> pprint.pprint([str(dt) for dt in dts_at_845])
-        ['2016-03-12 13:45:00+00:00',
-         '2016-03-13 12:45:00+00:00',
-         '2016-03-14 12:45:00+00:00']
-
-        :param days: DatetimeIndex An index of dates (represented as midnight).
-        :param market_time: datetime.time The time to apply as an offset to each day in ``days``.
-        :param day_offset: int The number of days we want to offset @days by
-        :return: pd.Series of date with the time requested.
-        """
-        # Offset days without tz to avoid timezone issues.
-        days = pd.DatetimeIndex(days).tz_localize(None).to_series()
-
-        if isinstance(
-            market_time, str
-        ):  # if string, assume its a reference to saved market times
-            timedeltas = self._regular_market_timedeltas[market_time]
-            datetimes = days + timedeltas[0][1]
-            for cut_off, timedelta in timedeltas[1:]:
-                datetimes = datetimes.where(
-                    days < pd.Timestamp(cut_off), days + timedelta
-                )
-
-        else:  # otherwise, assume it is a datetime.time object
-            datetimes = days + self._tdelta(market_time, day_offset)
-
-        return datetimes.dt.tz_localize(self.tz).dt.tz_convert("UTC")
-
-    def _tryholidays(self, cal, s, e):
-        try:
-            return cal.holidays(s, e)
-        except ValueError:
-            return pd.DatetimeIndex([])
-
-    def _special_dates(self, calendars, ad_hoc_dates, start, end):
-        """
-        Union an iterable of pairs of the form (time, calendar)
-        and an iterable of pairs of the form (time, [dates])
-
-        (This is shared logic for computing special opens and special closes.)
-        """
-        indexes = [
-            self.days_at_time(self._tryholidays(calendar, start, end), time_)
-            for time_, calendar in calendars
-        ] + [self.days_at_time(dates, time_) for time_, dates in ad_hoc_dates]
-        if indexes:
-            dates = pd.concat(indexes).sort_index().drop_duplicates()
-            return dates.loc[start : end.replace(hour=23, minute=59, second=59)]
-
-        return pd.Series([], dtype="datetime64[ns, UTC]", index=pd.DatetimeIndex([]))
-
-    def special_dates(self, market_time, start_date, end_date, filter_holidays=True):
-        """
-        Calculate a datetimeindex that only contains the specail times of the requested market time.
-
-        :param market_time: market_time reference
-        :param start_date: first possible date of the index
-        :param end_date: last possible date of the index
-        :param filter_holidays: will filter days by self.valid_days, which can be useful when debugging
-
-        :return: schedule DatetimeIndex
-        """
-        start_date, end_date = self.clean_dates(start_date, end_date)
-        calendars = self.get_special_times(market_time)
-        ad_hoc = self.get_special_times_adhoc(market_time)
-        special = self._special_dates(calendars, ad_hoc, start_date, end_date)
-
-        if filter_holidays:
-            valid = self.valid_days(start_date, end_date, tz=None)
-            special = special[
-                special.index.isin(valid)
-            ]  # some sources of special times don't exclude holidays
-        return special
-
-    def schedule(
-        self,
-        start_date,
-        end_date,
-        tz="UTC",
-        start="market_open",
-        end="market_close",
-        force_special_times=True,
-        market_times=None,
-        interruptions=False,
-    ):
-        """
-        Generates the schedule DataFrame. The resulting DataFrame will have all the valid business days as the index
-        and columns for the requested market times. The columns can be determined either by setting a range (inclusive
-        on both sides), using `start` and `end`, or by passing a list to `market_times'. A range of market_times is
-        derived from a list of market_times that are available to the instance, which are sorted based on the current
-        regular time. See examples/usage.ipynb for demonstrations.
-
-        All time zones are set to UTC by default. Setting the tz parameter will convert the columns to the desired
-        timezone, such as 'America/New_York'.
-
-        :param start_date: first date of the schedule
-        :param end_date: last date of the schedule
-        :param tz: timezone that the columns of the returned schedule are in, default: "UTC"
-        :param start: the first market_time to include as a column, default: "market_open"
-        :param end: the last market_time to include as a column, default: "market_close"
-        :param force_special_times: how to handle special times.
-            True: overwrite regular times of the column itself, conform other columns to special times of
-                market_open/market_close if those are requested.
-            False: only overwrite regular times of the column itself, leave others alone
-            None: completely ignore special times
-        :param market_times: alternative to start/end, list of market_times that are in self.regular_market_times
-        :param interruptions: bool, whether to add interruptions to the schedule, default: False
-            These will be added as columns to the right of the DataFrame. Any interruption on a day between
-            start_date and end_date will be included, regardless of the market_times requested.
-            Also, `force_special_times` does not take these into consideration.
-        :return: schedule DataFrame
-        """
-        start_date, end_date = self.clean_dates(start_date, end_date)
-        if not (start_date <= end_date):
-            raise ValueError("start_date must be before or equal to end_date.")
-
-        # Setup all valid trading days and the requested market_times
-        _all_days = self.valid_days(start_date, end_date)
-        if market_times is None:
-            market_times = self._get_market_times(start, end)
-        elif market_times == "all":
-            market_times = self._market_times
-
-        # If no valid days return an empty DataFrame
-        if not _all_days.size:
-            return pd.DataFrame(
-                columns=market_times, index=pd.DatetimeIndex([], freq="C")
-            )
-
-        _adj_others = force_special_times is True
-        _adj_col = force_special_times is not None
-        _open_adj = _close_adj = []
-
-        schedule = pd.DataFrame()
-        for market_time in market_times:
-            temp = self.days_at_time(_all_days, market_time).copy()  # standard times
-            if _adj_col:
-                # create an array of special times
-                special = self.special_dates(
-                    market_time, start_date, end_date, filter_holidays=False
-                )
-                # overwrite standard times
-                specialix = special.index[
-                    special.index.isin(temp.index)
-                ]  # some sources of special times don't exclude holidays
-                temp.loc[specialix] = special
-
-                if _adj_others:
-                    if market_time == "market_open":
-                        _open_adj = specialix
-                    elif market_time == "market_close":
-                        _close_adj = specialix
-
-            schedule[market_time] = temp
-
-        if _adj_others:
-            adjusted = schedule.loc[_open_adj].apply(
-                lambda x: x.where(x.ge(x["market_open"]), x["market_open"]), axis=1
-            )
-            schedule.loc[_open_adj] = adjusted
-
-            adjusted = schedule.loc[_close_adj].apply(
-                lambda x: x.where(x.le(x["market_close"]), x["market_close"]), axis=1
-            )
-            schedule.loc[_close_adj] = adjusted
-
-        if interruptions:
-            interrs = self.interruptions_df
-            schedule[interrs.columns] = interrs
-            schedule = schedule.dropna(how="all", axis=1)
-
-        if tz != "UTC":
-            schedule = schedule.apply(lambda s: s.dt.tz_convert(tz))
-
-        return schedule
-
-    def open_at_time(self, schedule, timestamp, include_close=False, only_rth=False):
-        """
-        Determine if a given timestamp is during an open time for the market. If the timestamp is
-        before the first open time or after the last close time of `schedule`, a ValueError will be raised.
-
-        :param schedule: schedule DataFrame
-        :param timestamp: the timestamp to check for. Assumed to be UTC, if it doesn't include tz information.
-        :param include_close: if False then the timestamp that equals the closing timestamp will return False and not be
-            considered a valid open date and time. If True then it will be considered valid and return True. Use True
-            if using bars and would like to include the last bar as a valid open date and time. The close refers to the
-            latest market_time available, which could be after market_close (e.g. 'post').
-        :param only_rth: whether to ignore columns that are before market_open or after market_close. If true,
-            include_close will be referring to market_close.
-        :return: True if the timestamp is a valid open date and time, False if not
-        """
-        timestamp = pd.Timestamp(timestamp)
-        try:
-            timestamp = timestamp.tz_localize("UTC")
-        except TypeError:
-            pass
-
-        cols = schedule.columns
-        interrs = cols.str.startswith("interruption_")
-        if not (cols.isin(self._oc_market_times) | interrs).all():
-            raise ValueError(
-                "You seem to be using a schedule that isn't based on the market_times, "
-                "or includes market_times that are not represented in the open_close_map."
-            )
-
-        if only_rth:
-            lowest, highest = "market_open", "market_close"
-        else:
-            cols = cols[~interrs]
-            ix = cols.map(self._oc_market_times.index)
-            lowest, highest = cols[ix == ix.min()][0], cols[ix == ix.max()][0]
-
-        if timestamp < schedule[lowest].iat[0] or timestamp > schedule[highest].iat[-1]:
-            raise ValueError("The provided timestamp is not covered by the schedule")
-
-        day = schedule[schedule[lowest].le(timestamp)].iloc[-1].dropna().sort_values()
-        day = day.loc[lowest:highest]
-        day = day.index.to_series(index=day)
-
-        if interrs.any():
-            starts = day.str.startswith("interruption_start_")
-            ends = day.str.startswith("interruption_end_")
-            day.loc[starts] = False
-            day.loc[ends] = True
-
-        # When post follows market_close, market_close should not be considered a close
-        day.loc[day.eq("market_close") & day.shift(-1).eq("post")] = "market_open"
-        day = day.replace(self.open_close_map)
-
-        if include_close:
-            below = day.index < timestamp
-        else:
-            below = day.index <= timestamp
-        return bool(day[below].iat[-1])  # returns numpy.bool_ if not bool(...)
-
-    # need this to make is_open_now testable
-    @staticmethod
-    def _get_current_time():
-        return pd.Timestamp.now(tz="UTC")
-
-    def is_open_now(self, schedule, include_close=False, only_rth=False):
-        """
-        To determine if the current local system time (converted to UTC) is an open time for the market
-
-        :param schedule: schedule DataFrame
-        :param include_close: if False then the function will return False if the current local system time is equal to
-            the closing timestamp. If True then it will return True if the current local system time is equal to the
-            closing timestamp. Use True if using bars and would like to include the last bar as a valid open date
-            and time.
-        :param only_rth: whether to consider columns that are before market_open or after market_close
-
-        :return: True if the current local system time is a valid open date and time, False if not
-        """
-        current_time = MarketCalendar._get_current_time()
-        return self.open_at_time(
-            schedule, current_time, include_close=include_close, only_rth=only_rth
-        )
-
-    def clean_dates(self, start_date, end_date):
-        """
-        Strips the inputs of time and time zone information
-
-        :param start_date: start date
-        :param end_date: end date
-        :return: (start_date, end_date) with just date, no time and no time zone
-        """
-        start_date = pd.Timestamp(start_date).tz_localize(None).normalize()
-        end_date = pd.Timestamp(end_date).tz_localize(None).normalize()
-        return start_date, end_date
-
-    def is_different(self, col, diff=None):
-        if diff is None:
-            diff = pd.Series.ne
-        normal = self.days_at_time(col.index, col.name)
-        return diff(col.dt.tz_convert("UTC"), normal)
-
-    def early_closes(self, schedule):
-        """
-        Get a DataFrame of the dates that are an early close.
-
-        :param schedule: schedule DataFrame
-        :return: schedule DataFrame with rows that are early closes
-        """
-        return schedule[self.is_different(schedule["market_close"], pd.Series.lt)]
-
-    def late_opens(self, schedule):
-        """
-        Get a DataFrame of the dates that are an late opens.
-
-        :param schedule: schedule DataFrame
-        :return: schedule DataFrame with rows that are late opens
-        """
-        return schedule[self.is_different(schedule["market_open"], pd.Series.gt)]
-
-    def __getitem__(self, item):
-        if isinstance(item, (tuple, list)):
-            if item[1] == "all":
-                return self.get_time(item[0], all_times=True)
-            else:
-                return self.get_time_on(item[0], item[1])
-        else:
-            return self.get_time(item)
-
-    def __setitem__(self, key, value):
-        return self.add_time(key, value)
-
-    def __delitem__(self, key):
-        return self.remove_time(key)
+# Fork of zipline from Quantopian. Licensed under MIT, original licence below
+#
+# Copyright 2016 Quantopian, Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+import warnings
+from abc import ABCMeta, abstractmethod
+from datetime import time
+from typing import Literal, Union
+
+import pandas as pd
+from pandas.tseries.offsets import CustomBusinessDay
+
+from .class_registry import RegisteryMeta, ProtectedDict
+
+from . import calendar_utils as u
+
+MONDAY, TUESDAY, WEDNESDAY, THURSDAY, FRIDAY, SATURDAY, SUNDAY = range(7)
+
+WEEKMASK_ABBR = {
+    MONDAY: "Mon",
+    TUESDAY: "Tue",
+    WEDNESDAY: "Wed",
+    THURSDAY: "Thu",
+    FRIDAY: "Fri",
+    SATURDAY: "Sat",
+    SUNDAY: "Sun",
+}
+
+
+class DEFAULT:
+    pass
+
+
+class MarketCalendarMeta(ABCMeta, RegisteryMeta):
+    pass
+
+
+class MarketCalendar(metaclass=MarketCalendarMeta):
+    """
+    An MarketCalendar represents the timing information of a single market or exchange.
+    Unless otherwise noted all times are in UTC and use Pandas data structures.
+    """
+
+    regular_market_times = {
+        "market_open": ((None, time(0)),),
+        "market_close": ((None, time(23)),),
+    }
+
+    open_close_map = {
+        "market_open": True,
+        "market_close": False,
+        "break_start": False,
+        "break_end": True,
+        "pre": True,
+        "post": False,
+    }
+
+    @staticmethod
+    def _tdelta(t, day_offset=0):
+        try:
+            return pd.Timedelta(
+                days=day_offset, hours=t.hour, minutes=t.minute, seconds=t.second
+            )
+        except AttributeError:
+            t, day_offset = t
+            return pd.Timedelta(
+                days=day_offset, hours=t.hour, minutes=t.minute, seconds=t.second
+            )
+
+    @staticmethod
+    def _off(tple):
+        try:
+            return tple[2]
+        except IndexError:
+            return 0
+
+    @classmethod
+    def calendar_names(cls):
+        """All Market Calendar names and aliases that can be used in "factory"
+        :return: list(str)
+        """
+        return [
+            cal
+            for cal in cls._regmeta_class_registry.keys()
+            if cal not in ["MarketCalendar", "TradingCalendar"]
+        ]
+
+    @classmethod
+    def factory(
+        cls, name, *args, **kwargs
+    ):  # Will be set by Meta, keeping it there for tests
+        """
+        :param name: The name of the MarketCalendar to be retrieved.
+        :param *args/**kwargs: passed to requested MarketCalendar.__init__
+        :return: MarketCalendar of the desired calendar.
+        """
+        return
+
+    def __init__(self, open_time=None, close_time=None):
+        """
+        :param open_time: Market open time override as datetime.time object. If None then default is used.
+        :param close_time: Market close time override as datetime.time object. If None then default is used.
+        """
+
+        self.regular_market_times = self.regular_market_times.copy()
+        self.open_close_map = self.open_close_map.copy()
+        self._customized_market_times = []
+
+        if open_time is not None:
+            self.change_time("market_open", open_time)
+
+        if close_time is not None:
+            self.change_time("market_close", close_time)
+
+        if not hasattr(self, "_market_times"):
+            self._prepare_regular_market_times()
+
+    @property
+    @abstractmethod
+    def name(self):
+        """
+        Name of the market
+
+        :return: string name
+        """
+        raise NotImplementedError()
+
+    @property
+    @abstractmethod
+    def tz(self):
+        """
+        Time zone for the market.
+
+        :return: timezone
+        """
+        raise NotImplementedError()
+
+    @property
+    def market_times(self):
+        return self._market_times
+
+    def _prepare_regular_market_times(self):
+        oc_map = self.open_close_map
+        assert all(
+            isinstance(x, bool) for x in oc_map.values()
+        ), "Values in open_close_map need to be True or False"
+
+        regular = self.regular_market_times
+        discontinued = ProtectedDict()
+        regular_tds = {}
+
+        for market_time, times in regular.items():
+            # in case a market_time has been discontinued, extend the last time
+            # and add it to the discontinued_market_times dictionary
+            if market_time.startswith("interruption_"):
+                raise ValueError("'interruption_' prefix is reserved")
+
+            if times[-1][1] is None:
+                discontinued._set(market_time, times[-1][0])
+                times = times[:-1]
+                regular._set(market_time, times)
+
+            regular_tds[market_time] = tuple(
+                (t[0], self._tdelta(t[1], self._off(t))) for t in times
+            )
+
+        if discontinued:
+            warnings.warn(
+                f"{list(discontinued.keys())} are discontinued, the dictionary"
+                f" `.discontinued_market_times` has the dates on which these were discontinued."
+                f" The times as of those dates are incorrect, use .remove_time(market_time)"
+                f" to ignore a market_time."
+            )
+
+        self.discontinued_market_times = discontinued
+        self.regular_market_times = regular
+
+        self._regular_market_timedeltas = regular_tds
+        self._market_times = sorted(regular.keys(), key=lambda x: regular_tds[x][-1][1])
+        self._oc_market_times = list(filter(oc_map.__contains__, self._market_times))
+
+    def _set_time(self, market_time, times, opens):
+        if isinstance(times, (tuple, list)):  # passed a tuple
+            if not isinstance(times[0], (tuple, list)):  # doesn't have a tuple inside
+                if times[0] is None:  # seems to be a tuple indicating starting time
+                    times = (times,)
+                else:  # must be a tuple with: (time, offset)
+                    times = ((None, times[0], times[1]),)
+        else:  # should be a datetime.time object
+            times = ((None, times),)
+
+        ln = len(times)
+        for i, t in enumerate(times):
+            try:
+                assert (
+                    t[0] is None
+                    or isinstance(t[0], str)
+                    or isinstance(t[0], pd.Timestamp)
+                )
+                assert isinstance(t[1], time) or (
+                    ln > 1 and i == ln - 1 and t[1] is None
+                )
+                assert isinstance(self._off(t), int)
+            except AssertionError:
+                raise AssertionError(
+                    "The passed time information is not in the right format, "
+                    "please consult the docs for how to set market times"
+                )
+
+        if opens is DEFAULT:
+            opens = self.__class__.open_close_map.get(market_time, None)
+
+        if opens in (True, False):
+            self.open_close_map._set(market_time, opens)
+
+        elif opens is None:  # make sure it's ignored
+            try:
+                self.open_close_map._del(market_time)
+            except KeyError:
+                pass
+        else:
+            raise ValueError(
+                "when you pass `opens`, it needs to be True, False, or None"
+            )
+
+        self.regular_market_times._set(market_time, times)
+
+        if not self.is_custom(market_time):
+            self._customized_market_times.append(market_time)
+
+        self._prepare_regular_market_times()
+
+    def change_time(self, market_time, times, opens=DEFAULT):
+        """
+        Changes the specified market time in regular_market_times and makes the necessary adjustments.
+
+        :param market_time: the market_time to change
+        :param times: new time information
+        :param opens: whether the market_time is a time that closes or opens the market
+            this is only needed if the market_time should be respected by .open_at_time
+            True: opens
+            False: closes
+            None: consider it neither opening nor closing, don't add to open_close_map (ignore in .open_at_time)
+            DEFAULT: same as None, unless the market_time is in self.__class__.open_close_map. Then it will take
+                the default value as defined by the class.
+        :return: None
+        """
+        assert market_time in self.regular_market_times, (
+            f"{market_time} is not in regular_market_times:" f"\n{self._market_times}."
+        )
+        return self._set_time(market_time, times, opens)
+
+    def add_time(self, market_time, times, opens=DEFAULT):
+        """
+        Adds the specified market time to regular_market_times and makes the necessary adjustments.
+
+        :param market_time: the market_time to add
+        :param times: the time information
+        :param opens: see .change_time docstring
+        :return: None
+        """
+        assert market_time not in self.regular_market_times, (
+            f"{market_time} is already in regular_market_times:"
+            f"\n{self._market_times}"
+        )
+
+        return self._set_time(market_time, times, opens)
+
+    def remove_time(self, market_time):
+        """
+        Removes the specified market time from regular_market_times and makes the necessary adjustments.
+
+        :param market_time: the market_time to remove
+        :return: None
+        """
+
+        self.regular_market_times._del(market_time)
+        try:
+            self.open_close_map._del(market_time)
+        except KeyError:
+            pass
+
+        self._prepare_regular_market_times()
+        if self.is_custom(market_time):
+            self._customized_market_times.remove(market_time)
+
+    def is_custom(self, market_time):
+        return market_time in self._customized_market_times
+
+    @property
+    def has_custom(self):
+        return len(self._customized_market_times) > 0
+
+    def is_discontinued(self, market_time):
+        return market_time in self.discontinued_market_times
+
+    @property
+    def has_discontinued(self):
+        return len(self.discontinued_market_times) > 0
+
+    def get_time(self, market_time, all_times=False):
+        try:
+            times = self.regular_market_times[market_time]
+        except KeyError as e:
+            if "break_start" in market_time or "break_end" in market_time:
+                return None  # in case of no breaks
+            elif market_time in ["market_open", "market_close"]:
+                raise NotImplementedError("You need to set market_times")
+            else:
+                raise e
+
+        if all_times:
+            return times
+        return times[-1][1].replace(tzinfo=self.tz)
+
+    def get_time_on(self, market_time, date):
+        times = self.get_time(market_time, all_times=True)
+        if times is None:
+            return None
+
+        date = pd.Timestamp(date)
+        for d, t in times[::-1]:
+            if d is None or pd.Timestamp(d) < date:
+                return t.replace(tzinfo=self.tz)
+
+    def open_time_on(self, date):
+        return self.get_time_on("market_open", date)
+
+    def close_time_on(self, date):
+        return self.get_time_on("market_close", date)
+
+    def break_start_on(self, date):
+        return self.get_time_on("break_start", date)
+
+    def break_end_on(self, date):
+        return self.get_time_on("break_end", date)
+
+    @property
+    def open_time(self):
+        """
+        Default open time for the market
+
+        :return: time
+        """
+        return self.get_time("market_open")
+
+    @property
+    def close_time(self):
+        """
+        Default close time for the market
+
+        :return: time
+        """
+        return self.get_time("market_close")
+
+    @property
+    def break_start(self):
+        """
+        Break time start. If None then there is no break
+
+        :return: time or None
+        """
+        return self.get_time("break_start")
+
+    @property
+    def break_end(self):
+        """
+        Break time end. If None then there is no break
+
+        :return: time or None
+        """
+        return self.get_time("break_end")
+
+    @property
+    def regular_holidays(self):
+        """
+
+        :return: pd.AbstractHolidayCalendar: a calendar containing the regular holidays for this calendar
+        """
+        return None
+
+    @property
+    def adhoc_holidays(self):
+        """
+
+        :return: list of ad-hoc holidays
+        """
+        return []
+
+    @property
+    def weekmask(self):
+        return "Mon Tue Wed Thu Fri"
+
+    @property
+    def special_opens(self):
+        """
+        A list of special open times and corresponding AbstractHolidayCalendar.
+
+        :return: List of (time, AbstractHolidayCalendar) tuples
+        """
+        return []
+
+    @property
+    def special_opens_adhoc(self):
+        """
+
+        :return: List of (time, DatetimeIndex) tuples that represent special opens that cannot be codified into rules.
+        """
+        return []
+
+    @property
+    def special_closes(self):
+        """
+        A list of special close times and corresponding HolidayCalendars.
+
+        :return: List of (time, AbstractHolidayCalendar) tuples
+        """
+        return []
+
+    @property
+    def special_closes_adhoc(self):
+        """
+
+        :return: List of (time, DatetimeIndex) tuples that represent special closes that cannot be codified into rules.
+        """
+        return []
+
+    def get_special_times(self, market_time):
+        return getattr(self, "special_" + market_time, [])
+
+    def get_special_times_adhoc(self, market_time):
+        return getattr(self, "special_" + market_time + "_adhoc", [])
+
+    def get_offset(self, market_time):
+        return self._off(self.get_time(market_time, all_times=True)[-1])
+
+    @property
+    def open_offset(self):
+        """
+        :return: open offset
+        """
+        return self.get_offset("market_open")
+
+    @property
+    def close_offset(self):
+        """
+        :return: close offset
+        """
+        return self.get_offset("market_close")
+
+    @property
+    def interruptions(self):
+        """
+        This needs to be a list with a tuple for each date that had an interruption.
+        The tuple should have this layout:
+
+            (date, start_time, end_time[, start_time2, end_time2, ...])
+
+        E.g.:
+        [
+            ("2002-02-03", (time(11), -1), time(11, 2)),
+            ("2010-01-11", time(11), (time(11, 1), 1)),
+            ("2010-01-13", time(9, 59), time(10), time(10, 29), time(10, 30)),
+            ("2011-01-10", time(11), time(11, 1))
+        ]
+
+        The date needs to be a string in this format: 'yyyy-mm-dd'.
+        Times need to be two datetime.time objects for each interruption, indicating start and end.
+         Optionally these can be wrapped in a tuple, where the
+         second element needs to be an integer indicating an offset.
+        On "2010-01-13" in the example, it is shown that there can be multiple interruptions in a day.
+        """
+        return []
+
+    def _convert(self, col: pd.Series):
+        """
+        col is a series indexed by dates at which interruptions occurred. The values are either the start or end times
+        of an interruption, represented by either a timedelta or a tuple with a timedelta and day offset of the form
+        (timedelta, offset). _convert produces a new series where the values are replaced by datetimes equal to the
+        index of the original series plus the offset if present, at the timedelta.
+
+        E.g.:
+        >>> self._convert(
+                pd.Series(
+                    [datetime.time(11, 2), (datetime.time(11, 1), 1), datetime.time(10, 0), None],
+                    index=pd.DatetimeIndex(['2002-02-03', '2010-01-11', '2010-01-13', '2011-01-10'])
+                )
+            )
+        2002-02-03   2002-02-03 11:02:00+00:00
+        2010-01-11   2010-01-12 11:01:00+00:00
+        2010-01-13   2010-01-13 10:00:00+00:00
+        2011-01-10                         NaT
+        dtype: datetime64[ns, UTC]
+        """
+        col = (
+            col.dropna()
+        )  # Python 3.8, pandas 2.0.3 cannot create time deltas from NaT
+        try:
+            times = col.str[0]
+        except AttributeError:  # no tuples, only offset 0
+            return (
+                (
+                    pd.to_timedelta(col.astype("string").fillna(""), errors="coerce")
+                    + col.index
+                )
+                .dt.tz_localize(self.tz)
+                .dt.tz_convert("UTC")
+            )
+
+        return (
+            (
+                pd.to_timedelta(
+                    times.fillna(col).astype("string").fillna(""), errors="coerce"
+                )
+                + pd.to_timedelta(col.str[1].fillna(0), unit="D")
+                + col.index
+            )
+            .dt.tz_localize(self.tz)
+            .dt.tz_convert("UTC")
+        )
+
+    @staticmethod
+    def _col_name(n: int):
+        return (
+            f"interruption_start_{n // 2 + 1}"
+            if n % 2 == 1
+            else f"interruption_end_{n // 2}"
+        )
+
+    @property
+    def interruptions_df(self):
+        """
+        Will return a pd.DataFrame only containing interruptions.
+        """
+        if not self.interruptions:
+            return pd.DataFrame(index=pd.DatetimeIndex([]))
+        intr = pd.DataFrame(self.interruptions)
+        intr.index = pd.to_datetime(intr.pop(0))
+
+        intr.columns = map(self._col_name, intr.columns)
+        intr.index.name = None
+
+        return intr.apply(self._convert).sort_index()
+
+    def holidays(self) -> pd.tseries.offsets.CustomBusinessDay:
+        """
+        Returns the complete CustomBusinessDay object of holidays that can be used in any Pandas function that take
+        that input.
+
+        :return: CustomBusinessDay object of holidays
+        """
+        try:
+            return self._holidays
+        except AttributeError:
+            self._holidays = CustomBusinessDay(
+                holidays=self.adhoc_holidays,
+                calendar=self.regular_holidays,
+                weekmask=self.weekmask,
+            )
+        return self._holidays
+
+    def valid_days(self, start_date, end_date, tz="UTC") -> pd.DatetimeIndex:
+        """
+        Get a DatetimeIndex of valid open business days.
+
+        :param start_date: start date
+        :param end_date: end date
+        :param tz: time zone in either string or pytz.timezone
+        :return: DatetimeIndex of valid business days
+        """
+        return pd.date_range(
+            start_date, end_date, freq=self.holidays(), normalize=True, tz=tz
+        )
+
+    def _get_market_times(self, start, end):
+        mts = self._market_times
+        return mts[mts.index(start) : mts.index(end) + 1]
+
+    def days_at_time(self, days, market_time, day_offset=0):
+        """
+        Create an index of days at time ``t``, interpreted in timezone ``tz``. The returned index is localized to UTC.
+
+        In the example below, the times switch from 13:45 to 12:45 UTC because
+        March 13th is the daylight savings transition for US/Eastern.  All the
+        times are still 8:45 when interpreted in US/Eastern.
+
+        >>> import pandas as pd; import datetime; import pprint
+        >>> dts = pd.date_range('2016-03-12', '2016-03-14')
+        >>> dts_at_845 = days_at_time(dts, datetime.time(8, 45), 'US/Eastern')
+        >>> pprint.pprint([str(dt) for dt in dts_at_845])
+        ['2016-03-12 13:45:00+00:00',
+         '2016-03-13 12:45:00+00:00',
+         '2016-03-14 12:45:00+00:00']
+
+        :param days: DatetimeIndex An index of dates (represented as midnight).
+        :param market_time: datetime.time The time to apply as an offset to each day in ``days``.
+        :param day_offset: int The number of days we want to offset @days by
+        :return: pd.Series of date with the time requested.
+        """
+        # Offset days without tz to avoid timezone issues.
+        days = pd.DatetimeIndex(days).tz_localize(None).to_series()
+
+        if isinstance(
+            market_time, str
+        ):  # if string, assume its a reference to saved market times
+            timedeltas = self._regular_market_timedeltas[market_time]
+            datetimes = days + timedeltas[0][1]
+            for cut_off, timedelta in timedeltas[1:]:
+                datetimes = datetimes.where(
+                    days < pd.Timestamp(cut_off), days + timedelta
+                )
+
+        else:  # otherwise, assume it is a datetime.time object
+            datetimes = days + self._tdelta(market_time, day_offset)
+
+        return datetimes.dt.tz_localize(self.tz).dt.tz_convert("UTC")
+
+    def _tryholidays(self, cal, s, e):
+        try:
+            # If the Calendar is all single Observance Holidays then it is far
+            # more efficient to extract and return those dates
+            observed_dates = u.all_single_observance_rules(cal)
+            if observed_dates is not None:
+                return pd.DatetimeIndex(
+                    [date for date in observed_dates if s <= date <= e]
+                )
+            else:
+                return cal.holidays(s, e)
+        except ValueError:
+            return pd.DatetimeIndex([])
+
+    def _special_dates(self, calendars, ad_hoc_dates, start, end):
+        """
+        Union an iterable of pairs of the forms (time, calendar),
+        (time, [dates]), and (time, int). If the second item in the pair
+        is an int it will be interpreted as a specific day of the week.
+
+        (This is shared logic for computing special opens and special closes.)
+        """
+        indexes = []
+        for time_, calendar in calendars:
+            if isinstance(calendar, int):
+                day_of_week = CustomBusinessDay(weekmask=WEEKMASK_ABBR[calendar])
+                indexes.append(
+                    self.days_at_time(
+                        pd.date_range(start, end, freq=day_of_week), time_
+                    )
+                )
+            else:
+                indexes.append(
+                    self.days_at_time(self._tryholidays(calendar, start, end), time_)
+                )
+
+        indexes += [self.days_at_time(dates, time_) for time_, dates in ad_hoc_dates]
+
+        if indexes:
+            dates = pd.concat(indexes).sort_index().drop_duplicates()
+            return dates.loc[start : end.replace(hour=23, minute=59, second=59)]
+
+        return pd.Series([], dtype="datetime64[ns, UTC]", index=pd.DatetimeIndex([]))
+
+    def special_dates(self, market_time, start_date, end_date, filter_holidays=True):
+        """
+        Calculate a datetimeindex that only contains the specail times of the requested market time.
+
+        :param market_time: market_time reference
+        :param start_date: first possible date of the index
+        :param end_date: last possible date of the index
+        :param filter_holidays: will filter days by self.valid_days, which can be useful when debugging
+
+        :return: schedule DatetimeIndex
+        """
+        start_date, end_date = self.clean_dates(start_date, end_date)
+        calendars = self.get_special_times(market_time)
+        ad_hoc = self.get_special_times_adhoc(market_time)
+        special = self._special_dates(calendars, ad_hoc, start_date, end_date)
+
+        if filter_holidays:
+            valid = self.valid_days(start_date, end_date, tz=None)
+            special = special[
+                special.index.isin(valid)
+            ]  # some sources of special times don't exclude holidays
+        return special
+
+    def schedule(
+        self,
+        start_date,
+        end_date,
+        tz="UTC",
+        start="market_open",
+        end="market_close",
+        force_special_times=True,
+        market_times=None,
+        interruptions=False,
+    ) -> pd.DataFrame:
+        """
+        Generates the schedule DataFrame. The resulting DataFrame will have all the valid business days as the index
+        and columns for the requested market times. The columns can be determined either by setting a range (inclusive
+        on both sides), using `start` and `end`, or by passing a list to `market_times'. A range of market_times is
+        derived from a list of market_times that are available to the instance, which are sorted based on the current
+        regular time. See examples/usage.ipynb for demonstrations.
+
+        All time zones are set to UTC by default. Setting the tz parameter will convert the columns to the desired
+        timezone, such as 'America/New_York'.
+
+        :param start_date: first date of the schedule
+        :param end_date: last date of the schedule
+        :param tz: timezone that the columns of the returned schedule are in, default: "UTC"
+        :param start: the first market_time to include as a column, default: "market_open"
+        :param end: the last market_time to include as a column, default: "market_close"
+        :param force_special_times: how to handle special times.
+            True: overwrite regular times of the column itself, conform other columns to special times of
+                market_open/market_close if those are requested.
+            False: only overwrite regular times of the column itself, leave others alone
+            None: completely ignore special times
+        :param market_times: alternative to start/end, list of market_times that are in self.regular_market_times
+        :param interruptions: bool, whether to add interruptions to the schedule, default: False
+            These will be added as columns to the right of the DataFrame. Any interruption on a day between
+            start_date and end_date will be included, regardless of the market_times requested.
+            Also, `force_special_times` does not take these into consideration.
+        :return: schedule DataFrame
+        """
+        start_date, end_date = self.clean_dates(start_date, end_date)
+        if not (start_date <= end_date):
+            raise ValueError("start_date must be before or equal to end_date.")
+
+        _all_days = self.valid_days(start_date, end_date)
+
+        # Setup all valid trading days and the requested market_times
+        if market_times is None:
+            market_times = self._get_market_times(start, end)
+        elif market_times == "all":
+            market_times = self._market_times
+
+        if not _all_days.size:  # If no valid days return an empty DataFrame
+            return pd.DataFrame(
+                columns=market_times, index=pd.DatetimeIndex([], freq="C")
+            )
+
+        return self.schedule_from_days(
+            _all_days, tz, start, end, force_special_times, market_times, interruptions
+        )
+
+    def schedule_from_days(
+        self,
+        days: pd.DatetimeIndex,
+        tz="UTC",
+        start="market_open",
+        end="market_close",
+        force_special_times=True,
+        market_times=None,
+        interruptions=False,
+    ) -> pd.DataFrame:
+        """
+        Generates a schedule DataFrame for the days provided. The days are assumed to be valid trading days.
+
+        The columns can be determined either by setting a range (inclusive on both sides), using `start` and `end`,
+        or by passing a list to `market_times'. A range of market_times is derived from a list of market_times that
+        are available to the instance, which are sorted based on the current regular time.
+        See examples/usage.ipynb for demonstrations.
+
+        All time zones are set to UTC by default. Setting the tz parameter will convert the columns to the desired
+        timezone, such as 'America/New_York'.
+
+        :param days: pd.DatetimeIndex of all the desired days in ascending order. This function does not double check
+            that these are valid trading days, it is assumed they are. It is intended that this parameter is generated
+            by either the .valid_days() or .date_range_htf() methods. Time & Timezone Information is ignored.
+        :param tz: timezone that the columns of the returned schedule are in, default: "UTC"
+        :param start: the first market_time to include as a column, default: "market_open"
+        :param end: the last market_time to include as a column, default: "market_close"
+        :param force_special_times: how to handle special times.
+            True: overwrite regular times of the column itself, conform other columns to special times of
+                market_open/market_close if those are requested.
+            False: only overwrite regular times of the column itself, leave others alone
+            None: completely ignore special times
+        :param market_times: alternative to start/end, list of market_times that are in self.regular_market_times
+        :param interruptions: bool, whether to add interruptions to the schedule, default: False
+            These will be added as columns to the right of the DataFrame. Any interruption on a day between
+            start_date and end_date will be included, regardless of the market_times requested.
+            Also, `force_special_times` does not take these into consideration.
+        :return: schedule DataFrame
+        """
+
+        if days.dtype != "datetime64[ns]":
+            days = pd.DatetimeIndex(days).normalize().tz_localize(None)
+
+        # Setup all valid trading days and the requested market_times
+        if market_times is None:
+            market_times = self._get_market_times(start, end)
+        elif market_times == "all":
+            market_times = self._market_times
+
+        _adj_others = force_special_times is True
+        _adj_col = force_special_times is not None
+        _open_adj = _close_adj = []
+
+        schedule = pd.DataFrame()
+        for market_time in market_times:
+            temp = self.days_at_time(days, market_time).copy()  # standard times
+            if _adj_col:
+                # create an array of special times
+                special = self.special_dates(
+                    market_time, days[0], days[-1], filter_holidays=False
+                )
+                # overwrite standard times
+                specialix = special.index[
+                    special.index.isin(temp.index)
+                ]  # some sources of special times don't exclude holidays
+                temp.loc[specialix] = special
+
+                if _adj_others:
+                    if market_time == "market_open":
+                        _open_adj = specialix
+                    elif market_time == "market_close":
+                        _close_adj = specialix
+
+            schedule[market_time] = temp
+
+        cols = schedule.columns
+        if _adj_others and len(_open_adj) > 0:
+            mkt_open_ind = cols.get_loc("market_open")
+
+            # Can't use Lambdas here since numpy array assignment doesn't return the array.
+            def adjust_opens(x):  # x is an np.Array.
+                x[x <= x[mkt_open_ind]] = x[mkt_open_ind]
+                return x
+
+            adjusted = schedule.loc[_open_adj].apply(adjust_opens, axis=1, raw=True)
+            schedule.loc[_open_adj] = adjusted
+
+        if _adj_others and len(_close_adj) > 0:
+            mkt_close_ind = cols.get_loc("market_close")
+
+            def adjust_closes(x):
+                x[x >= x[mkt_close_ind]] = x[mkt_close_ind]
+                return x
+
+            adjusted = schedule.loc[_close_adj].apply(adjust_closes, axis=1, raw=True)
+            schedule.loc[_close_adj] = adjusted
+
+        if interruptions:
+            interrs = self.interruptions_df
+            schedule[interrs.columns] = interrs
+            schedule = schedule.dropna(how="all", axis=1)
+
+        if tz != "UTC":
+            schedule = schedule.apply(lambda s: s.dt.tz_convert(tz))
+
+        return schedule
+
+    def date_range_htf(
+        self,
+        frequency: Union[str, pd.Timedelta, int, float],
+        start: Union[str, pd.Timestamp, int, float, None] = None,
+        end: Union[str, pd.Timestamp, int, float, None] = None,
+        periods: Union[int, None] = None,
+        closed: Union[Literal["left", "right"], None] = "right",
+        *,
+        day_anchor: u.Day_Anchor = "SUN",
+        month_anchor: u.Month_Anchor = "JAN",
+    ) -> pd.DatetimeIndex:
+        """
+        Returns a Normalized DatetimeIndex from the start-date to End-Date for Time periods of 1D and Higher.
+
+        PARAMETERS:
+
+        :param frequency: String, Int/float (POSIX seconds) or pd.Timedelta of the desired frequency.
+            :Must be Greater than '1D' and an integer multiple of the base frequency (D, W, M, Q, or Y)
+            :Important Note: Ints/Floats & Timedeltas are always considered as 'Open Business Days',
+                '2D' == Every Other Buisness Day, '3D' == Every 3rd B.Day, '7D' == Every 7th B.Day
+            :Higher periods (passed as strings) align to the beginning or end of the relevant period
+            :i.e. '1W' == First/[Last] Trading Day of each Week, '1Q' == First/[Last] Day of every Quarter
+
+        :param start: String, Int/float (POSIX seconds) or pd.Timestamp of the desired start time.
+            :The Time & Timezone information is ignored. Only the Normalized Day is considered.
+
+        :param end: String, Int/float (POSIX seconds) or pd.Timestamp of the desired start time.
+            :The Time & Timezone information is ignored. Only the Normalized Day is considered.
+
+        :param periods: Optional Integer number of periods to return. If a Period count, Start time,
+            and End time are given the period count is ignored.
+
+        :param closed: Literal['left', 'right']. Method used to close each range.
+            :Left: First open trading day of the Session is returned (e.g. First Open Day of The Month)
+            :right: Last open trading day of the Session is returned (e.g. Last Open Day of The Month)
+            :Note, This has no effect when the desired frequency is a number of days.
+
+        :param day_anchor: Day to Anchor the start of the Weekly timeframes to. Default 'SUN'.
+            : To get the First/Last Days of the trading Week then the Anchor needs to be on a day the relevant
+                market is closed.
+            : This can be set so that a specific day each week is returned.
+            : freq='1W' & day_anchor='WED' Will return Every 'WED' when the market is open, and nearest day
+                to the left or right (based on 'closed') when the market is closed.
+            Options: ["SUN", "MON", "TUE", "WED", "THU", "FRI", "SAT"]
+
+        :param month_anchor: Month to Anchor the start of the year to for Quarter and yearly timeframes.
+            : Default 'JAN' for Calendar Quarters/Years. Can be set to 'JUL' to return Fiscal Years
+            Options: ["JAN", "FEB", "MAR", "APR", "MAY", "JUN", "JUL", "AUG", "SEP", "OCT", "NOV", "DEC"]
+        """
+        return u.date_range_htf(
+            self.holidays(),
+            frequency,
+            start,
+            end,
+            periods,
+            closed,
+            day_anchor=day_anchor,
+            month_anchor=month_anchor,
+        )
+
+    def open_at_time(self, schedule, timestamp, include_close=False, only_rth=False):
+        """
+        Determine if a given timestamp is during an open time for the market. If the timestamp is
+        before the first open time or after the last close time of `schedule`, a ValueError will be raised.
+
+        :param schedule: schedule DataFrame
+        :param timestamp: the timestamp to check for. Assumed to be UTC, if it doesn't include tz information.
+        :param include_close: if False then the timestamp that equals the closing timestamp will return False and not be
+            considered a valid open date and time. If True then it will be considered valid and return True. Use True
+            if using bars and would like to include the last bar as a valid open date and time. The close refers to the
+            latest market_time available, which could be after market_close (e.g. 'post').
+        :param only_rth: whether to ignore columns that are before market_open or after market_close. If true,
+            include_close will be referring to market_close.
+        :return: True if the timestamp is a valid open date and time, False if not
+        """
+        timestamp = pd.Timestamp(timestamp)
+        try:
+            timestamp = timestamp.tz_localize("UTC")
+        except TypeError:
+            pass
+
+        cols = schedule.columns
+        interrs = cols.str.startswith("interruption_")
+        if not (cols.isin(self._oc_market_times) | interrs).all():
+            raise ValueError(
+                "You seem to be using a schedule that isn't based on the market_times, "
+                "or includes market_times that are not represented in the open_close_map."
+            )
+
+        if only_rth:
+            lowest, highest = "market_open", "market_close"
+        else:
+            cols = cols[~interrs]
+            ix = cols.map(self._oc_market_times.index)
+            lowest, highest = cols[ix == ix.min()][0], cols[ix == ix.max()][0]
+
+        if timestamp < schedule[lowest].iat[0] or timestamp > schedule[highest].iat[-1]:
+            raise ValueError("The provided timestamp is not covered by the schedule")
+
+        day = schedule[schedule[lowest].le(timestamp)].iloc[-1].dropna().sort_values()
+        day = day.loc[lowest:highest]
+        day = day.index.to_series(index=day)
+
+        if interrs.any():
+            starts = day.str.startswith("interruption_start_")
+            ends = day.str.startswith("interruption_end_")
+            day.loc[starts] = False
+            day.loc[ends] = True
+
+        # When post follows market_close, market_close should not be considered a close
+        day.loc[day.eq("market_close") & day.shift(-1).eq("post")] = "market_open"
+        day = day.map(
+            lambda x: (
+                self.open_close_map.get(x) if x in self.open_close_map.keys() else x
+            )
+        )
+
+        if include_close:
+            below = day.index < timestamp
+        else:
+            below = day.index <= timestamp
+        return bool(day[below].iat[-1])  # returns numpy.bool_ if not bool(...)
+
+    # need this to make is_open_now testable
+    @staticmethod
+    def _get_current_time():
+        return pd.Timestamp.now(tz="UTC")
+
+    def is_open_now(self, schedule, include_close=False, only_rth=False):
+        """
+        To determine if the current local system time (converted to UTC) is an open time for the market
+
+        :param schedule: schedule DataFrame
+        :param include_close: if False then the function will return False if the current local system time is equal to
+            the closing timestamp. If True then it will return True if the current local system time is equal to the
+            closing timestamp. Use True if using bars and would like to include the last bar as a valid open date
+            and time.
+        :param only_rth: whether to consider columns that are before market_open or after market_close
+
+        :return: True if the current local system time is a valid open date and time, False if not
+        """
+        current_time = MarketCalendar._get_current_time()
+        return self.open_at_time(
+            schedule, current_time, include_close=include_close, only_rth=only_rth
+        )
+
+    def clean_dates(self, start_date, end_date):
+        """
+        Strips the inputs of time and time zone information
+
+        :param start_date: start date
+        :param end_date: end date
+        :return: (start_date, end_date) with just date, no time and no time zone
+        """
+        start_date = pd.Timestamp(start_date).tz_localize(None).normalize()
+        end_date = pd.Timestamp(end_date).tz_localize(None).normalize()
+        return start_date, end_date
+
+    def is_different(self, col, diff=None):
+        if diff is None:
+            diff = pd.Series.ne
+        normal = self.days_at_time(col.index, col.name)
+        return diff(col.dt.tz_convert("UTC"), normal)
+
+    def early_closes(self, schedule):
+        """
+        Get a DataFrame of the dates that are an early close.
+
+        :param schedule: schedule DataFrame
+        :return: schedule DataFrame with rows that are early closes
+        """
+        return schedule[self.is_different(schedule["market_close"], pd.Series.lt)]
+
+    def late_opens(self, schedule):
+        """
+        Get a DataFrame of the dates that are an late opens.
+
+        :param schedule: schedule DataFrame
+        :return: schedule DataFrame with rows that are late opens
+        """
+        return schedule[self.is_different(schedule["market_open"], pd.Series.gt)]
+
+    def __getitem__(self, item):
+        if isinstance(item, (tuple, list)):
+            if item[1] == "all":
+                return self.get_time(item[0], all_times=True)
+            else:
+                return self.get_time_on(item[0], item[1])
+        else:
+            return self.get_time(item)
+
+    def __setitem__(self, key, value):
+        return self.add_time(key, value)
+
+    def __delitem__(self, key):
+        return self.remove_time(key)