wassname/catalyst
1
0
Fork 0
mirror of https://github.com/wassname/catalyst.git synced 2026-07-04 06:25:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #1455 from quantopian/cache-future-chain-lookups

Browse Source 
ENH: Put a cache in front of future chain lookups
This commit is contained in:
Jean Bredeche
2016-08-31 11:26:14 -04:00
committed by GitHub
parent 8863d007ce 8d61e16df9
commit f5d6bfca4d
1 changed files with 63 additions and 57 deletions
Download Patch File Download Diff File Expand all files Collapse all files
zipline/assets/assets.py
+63 -57
View File
@@ -1,4 +1,4 @@
# Copyright 2015 Quantopian, Inc.
# 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.
@@ -58,7 +58,7 @@ from .asset_db_schema import (
ASSET_DB_VERSION
)
from zipline.utils.control_flow import invert
from zipline.utils.memoize import lazyval
from zipline.utils.memoize import lazyval, weak_lru_cache
from zipline.utils.numpy_utils import as_column
from zipline.utils.preprocess import preprocess
from zipline.utils.sqlite_utils import group_into_chunks, coerce_string_to_eng
@@ -762,7 +762,6 @@ class AssetFinder(object):
Raised when a future chain could not be found for the given
root symbol.
"""
fc_cols = self.futures_contracts.c
if as_of_date is pd.NaT:
@@ -776,60 +775,10 @@ class AssetFinder(object):
fc_cols.notice_date.asc(),
).execute().fetchall()))
else:
as_of_date = as_of_date.value
sids = list(map(
itemgetter('sid'),
sa.select((fc_cols.sid,)).where(
(fc_cols.root_symbol == root_symbol) &
# Filter to contracts that are still valid. If both
# exist, use the one that comes first in time (i.e.
# the lower value). If either notice_date or
# expiration_date is NaT, use the other. If both are
# NaT, the contract cannot be included in any chain.
sa.case(
[
(
fc_cols.notice_date == pd.NaT.value,
fc_cols.expiration_date >= as_of_date
),
(
fc_cols.expiration_date == pd.NaT.value,
fc_cols.notice_date >= as_of_date
)
],
else_=(
sa.func.min(
fc_cols.notice_date,
fc_cols.expiration_date
) >= as_of_date
)
)
).order_by(
# If both dates exist sort using minimum of
# expiration_date and notice_date
# else if one is NaT use the other.
sa.case(
[
(
fc_cols.expiration_date == pd.NaT.value,
fc_cols.notice_date
),
(
fc_cols.notice_date == pd.NaT.value,
fc_cols.expiration_date
)
],
else_=(
sa.func.min(
fc_cols.notice_date,
fc_cols.expiration_date
)
)
).asc()
).execute().fetchall()
))
sids = self._get_future_sids_for_root_symbol(
root_symbol,
as_of_date.value
)
if not sids:
# Check if root symbol exists.
@@ -842,6 +791,63 @@ class AssetFinder(object):
contracts = self.retrieve_futures_contracts(sids)
return [contracts[sid] for sid in sids]
@weak_lru_cache(100)
def _get_future_sids_for_root_symbol(self, root_symbol, as_of_date_ns):
fc_cols = self.futures_contracts.c
return list(map(
itemgetter('sid'),
sa.select((fc_cols.sid,)).where(
(fc_cols.root_symbol == root_symbol) &
# Filter to contracts that are still valid. If both
# exist, use the one that comes first in time (i.e.
# the lower value). If either notice_date or
# expiration_date is NaT, use the other. If both are
# NaT, the contract cannot be included in any chain.
sa.case(
[
(
fc_cols.notice_date == pd.NaT.value,
fc_cols.expiration_date >= as_of_date_ns
),
(
fc_cols.expiration_date == pd.NaT.value,
fc_cols.notice_date >= as_of_date_ns
)
],
else_=(
sa.func.min(
fc_cols.notice_date,
fc_cols.expiration_date
) >= as_of_date_ns
)
)
).order_by(
# If both dates exist sort using minimum of
# expiration_date and notice_date
# else if one is NaT use the other.
sa.case(
[
(
fc_cols.expiration_date == pd.NaT.value,
fc_cols.notice_date
),
(
fc_cols.notice_date == pd.NaT.value,
fc_cols.expiration_date
)
],
else_=(
sa.func.min(
fc_cols.notice_date,
fc_cols.expiration_date
)
)
).asc()
).execute().fetchall()
))
def lookup_expired_futures(self, start, end):
if not isinstance(start, pd.Timestamp):
start = pd.Timestamp(start)