mirror of
https://github.com/wassname/catalyst.git
synced 2026-06-29 13:13:41 +08:00
Scott Sanderson
8e59d12daf
- Adds `zipline.pipeline.Pipeline`, a new user-facing class for managing pipelines of Modeling API expressions. - Adds `attach_pipeline` and `drain_pipeline` as API methods - Removes `add_factor` and `add_filter` as API methods. These have been replaced two new methods on `Pipeline`: `add`, and `apply_screen`. - Adding a `Filter` as a column no longer implicitly truncates rows from the Modelling API output. It simply causes a new column, of dtype `bool` to show up in the output. Removal of rows is now handled by the new `apply_screen` method of `Pipeline`. - Refactors the existing Modeling API tests to reflect the new APIs.
133 lines
3.9 KiB
Python
133 lines
3.9 KiB
Python
"""
|
|
Base class for FFC unit tests.
|
|
"""
|
|
from functools import wraps
|
|
from unittest import TestCase
|
|
|
|
from numpy import arange, prod
|
|
from pandas import date_range, Int64Index, DataFrame
|
|
from six import iteritems
|
|
|
|
from zipline.finance.trading import TradingEnvironment
|
|
from zipline.modelling.engine import SimpleFFCEngine
|
|
from zipline.modelling.term import AssetExists
|
|
from zipline.utils.pandas_utils import explode
|
|
from zipline.utils.test_utils import make_simple_asset_info, ExplodingObject
|
|
from zipline.utils.tradingcalendar import trading_day
|
|
|
|
|
|
def with_defaults(**default_funcs):
|
|
"""
|
|
Decorator for providing dynamic default values for a method.
|
|
|
|
Usages:
|
|
|
|
@with_defaults(foo=lambda self: self.x + self.y)
|
|
def func(self, foo):
|
|
...
|
|
|
|
If a value is passed for `foo`, it will be used. Otherwise the function
|
|
supplied to `with_defaults` will be called with `self` as an argument.
|
|
"""
|
|
def decorator(f):
|
|
@wraps(f)
|
|
def method(self, *args, **kwargs):
|
|
for name, func in iteritems(default_funcs):
|
|
if name not in kwargs:
|
|
kwargs[name] = func(self)
|
|
return f(self, *args, **kwargs)
|
|
return method
|
|
return decorator
|
|
|
|
|
|
with_default_shape = with_defaults(shape=lambda self: self.default_shape)
|
|
|
|
|
|
class BaseFFCTestCase(TestCase):
|
|
|
|
def setUp(self):
|
|
self.__calendar = date_range('2014', '2015', freq=trading_day)
|
|
self.__assets = assets = Int64Index(arange(1, 20))
|
|
|
|
# Set up env for test
|
|
env = TradingEnvironment()
|
|
env.write_data(
|
|
equities_df=make_simple_asset_info(
|
|
assets,
|
|
self.__calendar[0],
|
|
self.__calendar[-1],
|
|
),
|
|
)
|
|
self.__finder = env.asset_finder
|
|
|
|
# Use a 30-day period at the end of the year by default.
|
|
self.__mask = self.__finder.lifetimes(
|
|
self.__calendar[-30:],
|
|
include_start_date=False,
|
|
)
|
|
|
|
@property
|
|
def default_shape(self):
|
|
"""Default shape for methods that build test data."""
|
|
return self.__mask.shape
|
|
|
|
def run_graph(self, graph, initial_workspace, mask=None):
|
|
"""
|
|
Compute the given TermGraph, seeding the workspace of our engine with
|
|
`initial_workspace`.
|
|
|
|
Parameters
|
|
----------
|
|
graph : zipline.pipeline.graph.TermGraph
|
|
Graph to run.
|
|
initial_workspace : dict
|
|
Initial workspace to forward to SimpleFFCEngine.compute_chunk.
|
|
mask : DataFrame, optional
|
|
This is a value to pass to `initial_workspace` as the mask from
|
|
`AssetExists()`. Defaults to a frame of shape `self.default_shape`
|
|
containing all True values.
|
|
|
|
Returns
|
|
-------
|
|
results : dict
|
|
Mapping from termname -> computed result.
|
|
"""
|
|
engine = SimpleFFCEngine(
|
|
ExplodingObject(),
|
|
self.__calendar,
|
|
self.__finder,
|
|
)
|
|
if mask is None:
|
|
mask = self.__mask
|
|
|
|
dates, assets, mask_values = explode(mask)
|
|
initial_workspace.setdefault(AssetExists(), mask_values)
|
|
return engine.compute_chunk(
|
|
graph,
|
|
dates,
|
|
assets,
|
|
initial_workspace,
|
|
)
|
|
|
|
def build_mask(self, array):
|
|
"""
|
|
Helper for constructing an AssetExists mask from a boolean-coercible
|
|
array.
|
|
"""
|
|
ndates, nassets = array.shape
|
|
return DataFrame(
|
|
array,
|
|
# Use the **last** N dates rather than the first N so that we have
|
|
# space for lookbacks.
|
|
index=self.__calendar[-ndates:],
|
|
columns=self.__assets[:nassets],
|
|
dtype=bool,
|
|
)
|
|
|
|
@with_default_shape
|
|
def arange_data(self, shape, dtype=float):
|
|
"""
|
|
Build a block of testing data from numpy.arange.
|
|
"""
|
|
return arange(prod(shape), dtype=dtype).reshape(shape)
|