wassname/catalyst
1
0
Fork 0
mirror of https://github.com/wassname/catalyst.git synced 2026-07-08 21:01:50 +08:00
Code Issues Packages Projects Releases Wiki Activity

ENH: add column for currency type

Browse Source
This commit is contained in:
Maya Tydykov
2016-05-05 11:20:53 -04:00
parent 18d1577d56
commit 8a3b82c536
5 changed files with 128 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/pipeline/test_dividends.py
+53 -8
View File
@@ -15,6 +15,7 @@ from zipline.pipeline.common import (
NEXT_EX_DATE,
NEXT_PAY_DATE,
PREVIOUS_ANNOUNCEMENT,
PREVIOUS_CURRENCY_TYPE,
PREVIOUS_EX_DATE,
PREVIOUS_PAY_DATE,
PREVIOUS_AMOUNT,
@@ -22,8 +23,8 @@ from zipline.pipeline.common import (
TS_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME,
EX_DATE_FIELD_NAME,
PAY_DATE_FIELD_NAME
)
PAY_DATE_FIELD_NAME,
CURRENCY_FIELD_NAME, NEXT_CURRENCY_TYPE)
from zipline.pipeline.data.dividends import (
DividendsByAnnouncementDate,
DividendsByExDate,
@@ -46,7 +47,8 @@ from zipline.pipeline.loaders.dividends import (
)
from zipline.pipeline.loaders.utils import (
zip_with_dates,
zip_with_floats
zip_with_floats,
zip_with_strs,
)
from zipline.testing.fixtures import (
WithPipelineEventDataLoader,
@@ -60,7 +62,8 @@ dividends_cases = [
EX_DATE_FIELD_NAME: pd.to_datetime(['2014-01-15', '2014-01-20']),
PAY_DATE_FIELD_NAME: pd.to_datetime(['2014-01-15', '2014-01-20']),
TS_FIELD_NAME: pd.to_datetime(['2014-01-05', '2014-01-10']),
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-09'])
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-09']),
CURRENCY_FIELD_NAME: ["$", "EUR"]
}),
# K1--K2--A2--A1.
pd.DataFrame({
@@ -68,7 +71,8 @@ dividends_cases = [
EX_DATE_FIELD_NAME: pd.to_datetime(['2014-01-20', '2014-01-15']),
PAY_DATE_FIELD_NAME: pd.to_datetime(['2014-01-20', '2014-01-15']),
TS_FIELD_NAME: pd.to_datetime(['2014-01-05', '2014-01-10']),
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-09'])
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-09']),
CURRENCY_FIELD_NAME: ["EUR", "$"]
}),
# K1--A1--K2--A2.
pd.DataFrame({
@@ -76,7 +80,8 @@ dividends_cases = [
EX_DATE_FIELD_NAME: pd.to_datetime(['2014-01-10', '2014-01-20']),
PAY_DATE_FIELD_NAME: pd.to_datetime(['2014-01-10', '2014-01-20']),
TS_FIELD_NAME: pd.to_datetime(['2014-01-05', '2014-01-15']),
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-14'])
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-14']),
CURRENCY_FIELD_NAME: ["$", "EUR"]
}),
# K1 == K2.
pd.DataFrame({
@@ -84,14 +89,16 @@ dividends_cases = [
EX_DATE_FIELD_NAME: pd.to_datetime(['2014-01-10', '2014-01-15']),
PAY_DATE_FIELD_NAME: pd.to_datetime(['2014-01-10', '2014-01-15']),
TS_FIELD_NAME: pd.to_datetime(['2014-01-05'] * 2),
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-04'])
ANNOUNCEMENT_FIELD_NAME: pd.to_datetime(['2014-01-04', '2014-01-04']),
CURRENCY_FIELD_NAME: ["$", "EUR"]
}),
pd.DataFrame(
columns=[CASH_AMOUNT_FIELD_NAME,
EX_DATE_FIELD_NAME,
PAY_DATE_FIELD_NAME,
TS_FIELD_NAME,
ANNOUNCEMENT_FIELD_NAME],
ANNOUNCEMENT_FIELD_NAME,
CURRENCY_FIELD_NAME],
dtype='datetime64[ns]'
),
]
@@ -157,6 +164,16 @@ next_amounts = [['NaN', 1, 15, 'NaN'],
['NaN', 3, 'NaN', 1, 'NaN'],
['NaN', 6, 23, 'NaN']]
prev_currency_types = [[None, "$", "EUR"],
[None, "$", "EUR"],
[None, "$", "EUR"],
[None, "$", "EUR"]]
next_currency_types = [[None, "$", "EUR", None],
[None, "EUR", "$", "EUR", None],
[None, "$", None, "EUR", None],
[None, "$", "EUR", None]]
class DividendsByAnnouncementDateTestCase(WithPipelineEventDataLoader,
ZiplineTestCase):
@@ -169,6 +186,8 @@ class DividendsByAnnouncementDateTestCase(WithPipelineEventDataLoader,
PREVIOUS_AMOUNT: DividendsByAnnouncementDate.previous_amount.latest,
DAYS_SINCE_PREV_DIVIDEND_ANNOUNCEMENT:
BusinessDaysSinceDividendAnnouncement(),
PREVIOUS_CURRENCY_TYPE:
DividendsByAnnouncementDate.previous_currency.latest,
}
@classmethod
@@ -204,6 +223,8 @@ class DividendsByAnnouncementDateTestCase(WithPipelineEventDataLoader,
['NaT', '2014-01-04', '2014-01-14'],
['NaT', '2014-01-04']]
amounts = [['NaN', 1, 15], ['NaN', 7, 13], ['NaN', 3, 1], ['NaN', 23]]
currency_types = [[None, "$", "EUR"], [None, "$", "EUR"],
[None, "$", "EUR"]]
cols = {
PREVIOUS_ANNOUNCEMENT: self.get_sids_to_frames(
zip_with_dates, announcement_dates, date_intervals, dates,
@@ -212,6 +233,10 @@ class DividendsByAnnouncementDateTestCase(WithPipelineEventDataLoader,
PREVIOUS_AMOUNT: self.get_sids_to_frames(
zip_with_floats, amounts, date_intervals, dates, 'float', 'NaN'
),
PREVIOUS_CURRENCY_TYPE: self.get_sids_to_frames(
zip_with_strs, currency_types, date_intervals, dates,
'category', None
),
}
cols[
@@ -263,6 +288,8 @@ class DividendsByExDateTestCase(WithPipelineEventDataLoader, ZiplineTestCase):
PREVIOUS_EX_DATE: DividendsByExDate.previous_date.latest,
NEXT_AMOUNT: DividendsByExDate.next_amount.latest,
PREVIOUS_AMOUNT: DividendsByExDate.previous_amount.latest,
PREVIOUS_CURRENCY_TYPE: DividendsByExDate.previous_currency.latest,
NEXT_CURRENCY_TYPE: DividendsByExDate.next_currency.latest,
DAYS_TO_NEXT_EX_DATE: BusinessDaysUntilNextExDate(),
DAYS_SINCE_PREV_EX_DATE: BusinessDaysSincePreviousExDate()
}
@@ -296,6 +323,14 @@ class DividendsByExDateTestCase(WithPipelineEventDataLoader, ZiplineTestCase):
PREVIOUS_AMOUNT: self.get_sids_to_frames(
zip_with_floats, prev_amounts, prev_date_intervals, dates,
'float', 'NaN'
),
PREVIOUS_CURRENCY_TYPE: self.get_sids_to_frames(
zip_with_strs, prev_currency_types, prev_date_intervals, dates,
'category', None
),
NEXT_CURRENCY_TYPE: self.get_sids_to_frames(
zip_with_strs, next_currency_types, next_date_intervals, dates,
'category', None
)
}
@@ -349,6 +384,8 @@ class DividendsByPayDateTestCase(WithPipelineEventDataLoader, ZiplineTestCase):
PREVIOUS_PAY_DATE: DividendsByPayDate.previous_date.latest,
NEXT_AMOUNT: DividendsByPayDate.next_amount.latest,
PREVIOUS_AMOUNT: DividendsByPayDate.previous_amount.latest,
PREVIOUS_CURRENCY_TYPE: DividendsByPayDate.previous_currency.latest,
NEXT_CURRENCY_TYPE: DividendsByPayDate.next_currency.latest,
}
@classmethod
@@ -380,6 +417,14 @@ class DividendsByPayDateTestCase(WithPipelineEventDataLoader, ZiplineTestCase):
PREVIOUS_AMOUNT: self.get_sids_to_frames(
zip_with_floats, prev_amounts, prev_date_intervals, dates,
'float', 'NaN'
),
PREVIOUS_CURRENCY_TYPE: self.get_sids_to_frames(
zip_with_strs, prev_currency_types, prev_date_intervals, dates,
'category', None
),
NEXT_CURRENCY_TYPE: self.get_sids_to_frames(
zip_with_strs, next_currency_types, next_date_intervals, dates,
'category', None
)
}
zipline/pipeline/common.py
+10 -3
View File
@@ -5,6 +5,10 @@ ACTUAL_VALUE_FIELD_NAME = 'actual_value'
AD_FIELD_NAME = 'asof_date'
ANNOUNCEMENT_FIELD_NAME = 'announcement_date'
BUYBACK_AMOUNT_FIELD_NAME = 'buyback_amount'
CASH_FIELD_NAME = 'cash'
CASH_AMOUNT_FIELD_NAME = 'cash_amount'
COUNT_FIELD_NAME = 'count'
CURRENCY_FIELD_NAME = 'currency_type'
BUYBACK_ANNOUNCEMENT_FIELD_NAME = 'buyback_date'
BUYBACK_TYPE_FIELD_NAME = 'buyback_type'
BUYBACK_UNIT_FIELD_NAME = 'buyback_unit'
@@ -27,6 +31,7 @@ MEAN_FIELD_NAME = 'mean'
NEXT_AMOUNT = 'next_amount'
NEXT_ANNOUNCEMENT = 'next_announcement'
NEXT_COUNT = 'next_count'
NEXT_CURRENCY_TYPE = 'next_currency_type'
NEXT_EX_DATE = 'next_ex_date'
NEXT_FISCAL_QUARTER = 'next_fiscal_quarter'
NEXT_FISCAL_YEAR = 'next_fiscal_year'
@@ -46,10 +51,12 @@ PREVIOUS_BUYBACK_AMOUNT = 'previous_value'
PREVIOUS_BUYBACK_ANNOUNCEMENT = 'previous_buyback_announcement'
PREVIOUS_BUYBACK_CASH = 'previous_buyback_cash'
PREVIOUS_BUYBACK_SHARE_COUNT = 'previous_buyback_share_count'
PREVIOUS_BUYBACK_TYPE = 'previous_buyback_type'
PREVIOUS_BUYBACK_UNIT = 'previous_value_type'
PREVIOUS_COUNT = 'previous_count'
PREVIOUS_DISCLOSURE_DATE = 'previous_disclosure_date'
PREVIOUS_COUNT = 'previous_count'
PREVIOUS_BUYBACK_TYPE = 'previous_buyback_type'
PREVIOUS_VALUE = 'previous_value'
PREVIOUS_VALUE_TYPE = 'previous_value_type'
PREVIOUS_CURRENCY_TYPE = 'previous_currency_type'
PREVIOUS_EX_DATE = 'previous_ex_date'
PREVIOUS_FISCAL_QUARTER = 'previous_fiscal_quarter'
PREVIOUS_FISCAL_YEAR = 'previous_fiscal_year'
zipline/pipeline/data/dividends.py
+10 -1
View File
@@ -1,7 +1,11 @@
"""
Dataset representing dates of upcoming dividends.
"""
from zipline.utils.numpy_utils import datetime64ns_dtype, float64_dtype
from zipline.utils.numpy_utils import (
datetime64ns_dtype,
float64_dtype,
categorical_dtype
)
from .dataset import Column, DataSet
@@ -11,6 +15,8 @@ class DividendsByExDate(DataSet):
previous_date = Column(datetime64ns_dtype)
next_amount = Column(float64_dtype)
previous_amount = Column(float64_dtype)
next_currency = Column(categorical_dtype)
previous_currency = Column(categorical_dtype)
class DividendsByPayDate(DataSet):
@@ -18,8 +24,11 @@ class DividendsByPayDate(DataSet):
previous_date = Column(datetime64ns_dtype)
next_amount = Column(float64_dtype)
previous_amount = Column(float64_dtype)
next_currency = Column(categorical_dtype)
previous_currency = Column(categorical_dtype)
class DividendsByAnnouncementDate(DataSet):
previous_announcement_date = Column(datetime64ns_dtype)
previous_amount = Column(float64_dtype)
previous_currency = Column(categorical_dtype)
zipline/pipeline/loaders/blaze/dividends.py
+11 -1
View File
@@ -1,6 +1,7 @@
from zipline.pipeline.common import (
ANNOUNCEMENT_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME,
EX_DATE_FIELD_NAME,
PAY_DATE_FIELD_NAME,
SID_FIELD_NAME,
@@ -47,6 +48,7 @@ class BlazeDividendsByAnnouncementDateLoader(BlazeEventsLoader):
{TS_FIELD_NAME}: datetime,
{CASH_AMOUNT_FIELD_NAME}: ?float64,
{ANNOUNCEMENT_FIELD_NAME}: ?datetime,
{CURRENCY_FIELD_NAME}: ?string,
}}
Where each row of the table is a record including the sid to identify the
@@ -61,13 +63,15 @@ class BlazeDividendsByAnnouncementDateLoader(BlazeEventsLoader):
TS_FIELD_NAME=TS_FIELD_NAME,
SID_FIELD_NAME=SID_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME=CASH_AMOUNT_FIELD_NAME,
ANNOUNCEMENT_FIELD_NAME=ANNOUNCEMENT_FIELD_NAME
ANNOUNCEMENT_FIELD_NAME=ANNOUNCEMENT_FIELD_NAME,
CURRENCY_FIELD_NAME=CURRENCY_FIELD_NAME,
)
_expected_fields = frozenset({
TS_FIELD_NAME,
SID_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME,
ANNOUNCEMENT_FIELD_NAME
})
@@ -103,6 +107,7 @@ class BlazeDividendsByExDateLoader(BlazeEventsLoader):
{TS_FIELD_NAME}: datetime,
{EX_DATE_FIELD_NAME}: ?datetime,
{CASH_AMOUNT_FIELD_NAME}: ?datetime,
{CURRENCY_FIELD_NAME}: ?string,
}}
Where each row of the table is a record including the sid to identify the
@@ -118,6 +123,7 @@ class BlazeDividendsByExDateLoader(BlazeEventsLoader):
SID_FIELD_NAME=SID_FIELD_NAME,
EX_DATE_FIELD_NAME=EX_DATE_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME=CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME=CURRENCY_FIELD_NAME,
)
_expected_fields = frozenset({
@@ -125,6 +131,7 @@ class BlazeDividendsByExDateLoader(BlazeEventsLoader):
SID_FIELD_NAME,
EX_DATE_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME,
})
concrete_loader = DividendsByExDateLoader
@@ -159,6 +166,7 @@ class BlazeDividendsByPayDateLoader(BlazeEventsLoader):
{TS_FIELD_NAME}: datetime,
{PAY_DATE_FIELD_NAME}: ?datetime,
{CASH_AMOUNT_FIELD_NAME}: ?datetime,
{CURRENCY_FIELD_NAME}: ?string,
}}
Where each row of the table is a record including the sid to identify the
@@ -174,6 +182,7 @@ class BlazeDividendsByPayDateLoader(BlazeEventsLoader):
SID_FIELD_NAME=SID_FIELD_NAME,
PAY_DATE_FIELD_NAME=PAY_DATE_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME=CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME=CURRENCY_FIELD_NAME
)
_expected_fields = frozenset({
@@ -181,6 +190,7 @@ class BlazeDividendsByPayDateLoader(BlazeEventsLoader):
SID_FIELD_NAME,
PAY_DATE_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME,
})
concrete_loader = DividendsByPayDateLoader
zipline/pipeline/loaders/dividends.py
+44 -5
View File
@@ -1,8 +1,9 @@
from zipline.pipeline.common import (
ANNOUNCEMENT_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME,
EX_DATE_FIELD_NAME,
PAY_DATE_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME,
ANNOUNCEMENT_FIELD_NAME
)
from zipline.pipeline.loaders.events import EventsLoader
from ..data import (
@@ -15,7 +16,8 @@ from zipline.utils.memoize import lazyval
class DividendsByAnnouncementDateLoader(EventsLoader):
expected_cols = frozenset([ANNOUNCEMENT_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME])
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME])
event_date_col = ANNOUNCEMENT_FIELD_NAME
@@ -39,10 +41,18 @@ class DividendsByAnnouncementDateLoader(EventsLoader):
CASH_AMOUNT_FIELD_NAME
)
@lazyval
def previous_currency_loader(self):
return self._previous_event_value_loader(
self.dataset.previous_currency,
CURRENCY_FIELD_NAME
)
class DividendsByPayDateLoader(EventsLoader):
expected_cols = frozenset([PAY_DATE_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME])
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME])
event_date_col = PAY_DATE_FIELD_NAME
@@ -75,10 +85,25 @@ class DividendsByPayDateLoader(EventsLoader):
CASH_AMOUNT_FIELD_NAME
)
@lazyval
def previous_currency_loader(self):
return self._previous_event_value_loader(
self.dataset.previous_currency,
CURRENCY_FIELD_NAME
)
@lazyval
def next_currency_loader(self):
return self._next_event_value_loader(
self.dataset.next_currency,
CURRENCY_FIELD_NAME
)
class DividendsByExDateLoader(EventsLoader):
expected_cols = frozenset([EX_DATE_FIELD_NAME,
CASH_AMOUNT_FIELD_NAME])
CASH_AMOUNT_FIELD_NAME,
CURRENCY_FIELD_NAME])
event_date_col = EX_DATE_FIELD_NAME
@@ -110,3 +135,17 @@ class DividendsByExDateLoader(EventsLoader):
self.dataset.previous_amount,
CASH_AMOUNT_FIELD_NAME
)
@lazyval
def previous_currency_loader(self):
return self._previous_event_value_loader(
self.dataset.previous_currency,
CURRENCY_FIELD_NAME
)
@lazyval
def next_currency_loader(self):
return self._next_event_value_loader(
self.dataset.next_currency,
CURRENCY_FIELD_NAME
)