wassname/catalyst
1
0
Fork 0
mirror of https://github.com/wassname/catalyst.git synced 2026-07-10 00:33:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

ENH: Allows Float64Adjustments to act on a range of columns

Browse Source
This commit is contained in:
llllllllll
2015-10-13 18:08:11 -04:00
parent c7ca0166cc
commit 0183d0a914
5 changed files with 115 additions and 55 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/pipeline/test_adjusted_array.py
+12 -12
View File
@@ -95,7 +95,7 @@ def _gen_multiplicative_adjustment_cases(dtype):
# Note that row indices are inclusive!
adjustments[1] = [
adjustment_type(0, 0, 0, dtype(2)),
adjustment_type(0, 0, 0, 0, dtype(2)),
]
buffer_as_of[1] = array([[2, 1, 1],
[1, 1, 1],
@@ -108,8 +108,8 @@ def _gen_multiplicative_adjustment_cases(dtype):
buffer_as_of[2] = buffer_as_of[1]
adjustments[3] = [
adjustment_type(1, 2, 1, dtype(3)),
adjustment_type(0, 1, 0, dtype(4)),
adjustment_type(1, 2, 1, 1, dtype(3)),
adjustment_type(0, 1, 0, 0, dtype(4)),
]
buffer_as_of[3] = array([[8, 1, 1],
[4, 3, 1],
@@ -119,7 +119,7 @@ def _gen_multiplicative_adjustment_cases(dtype):
[1, 1, 1]], dtype=dtype)
adjustments[4] = [
adjustment_type(0, 3, 2, dtype(5))
adjustment_type(0, 3, 2, 2, dtype(5))
]
buffer_as_of[4] = array([[8, 1, 5],
[4, 3, 5],
@@ -129,8 +129,8 @@ def _gen_multiplicative_adjustment_cases(dtype):
[1, 1, 1]], dtype=dtype)
adjustments[5] = [
adjustment_type(0, 4, 1, dtype(6)),
adjustment_type(2, 2, 2, dtype(7)),
adjustment_type(0, 4, 1, 1, dtype(6)),
adjustment_type(2, 2, 2, 2, dtype(7)),
]
buffer_as_of[5] = array([[8, 6, 5],
[4, 18, 5],
@@ -162,7 +162,7 @@ def _gen_overwrite_adjustment_cases(dtype):
# Note that row indices are inclusive!
adjustments[1] = [
adjustment_type(0, 0, 0, dtype(1)),
adjustment_type(0, 0, 0, 0, dtype(1)),
]
buffer_as_of[1] = array([[1, 2, 2],
[2, 2, 2],
@@ -175,8 +175,8 @@ def _gen_overwrite_adjustment_cases(dtype):
buffer_as_of[2] = buffer_as_of[1]
adjustments[3] = [
adjustment_type(1, 2, 1, dtype(3)),
adjustment_type(0, 1, 0, dtype(4)),
adjustment_type(1, 2, 1, 1, dtype(3)),
adjustment_type(0, 1, 0, 0, dtype(4)),
]
buffer_as_of[3] = array([[4, 2, 2],
[4, 3, 2],
@@ -186,7 +186,7 @@ def _gen_overwrite_adjustment_cases(dtype):
[2, 2, 2]], dtype=dtype)
adjustments[4] = [
adjustment_type(0, 3, 2, dtype(5))
adjustment_type(0, 3, 2, 2, dtype(5))
]
buffer_as_of[4] = array([[4, 2, 5],
[4, 3, 5],
@@ -196,8 +196,8 @@ def _gen_overwrite_adjustment_cases(dtype):
[2, 2, 2]], dtype=dtype)
adjustments[5] = [
adjustment_type(0, 4, 1, dtype(6)),
adjustment_type(2, 2, 2, dtype(7)),
adjustment_type(0, 4, 1, 1, dtype(6)),
adjustment_type(2, 2, 2, 2, dtype(7)),
]
buffer_as_of[5] = array([[4, 6, 5],
[4, 6, 5],
tests/pipeline/test_frameload.py
+28 -4
View File
@@ -189,14 +189,38 @@ class DataFrameLoaderTestCase(TestCase):
)
expected_formatted_adjustments = {
6: [
Float64Multiply(first_row=0, last_row=5, col=0, value=0.5),
Float64Add(first_row=0, last_row=5, col=1, value=1.0),
Float64Multiply(
first_row=0,
last_row=5,
first_col=0,
last_col=0,
value=0.5,
),
Float64Add(
first_row=0,
last_row=5,
first_col=1,
last_col=1,
value=1.0,
),
],
7: [
Float64Add(first_row=5, last_row=6, col=1, value=1.0),
Float64Add(
first_row=5,
last_row=6,
first_col=1,
last_col=1,
value=1.0,
),
],
8: [
Float64Overwrite(first_row=6, last_row=7, col=2, value=99.0)
Float64Overwrite(
first_row=6,
last_row=7,
first_col=2,
last_col=2,
value=99.0,
)
],
}
self.assertEqual(formatted_adjustments, expected_formatted_adjustments)
tests/pipeline/test_us_equity_pricing_loader.py
+4 -2
View File
@@ -337,7 +337,8 @@ class USEquityPricingLoaderTestCase(TestCase):
Float64Multiply(
first_row=0,
last_row=delta,
col=sid - 1,
first_col=sid - 1,
last_col=sid - 1,
value=ratio,
)
)
@@ -347,7 +348,8 @@ class USEquityPricingLoaderTestCase(TestCase):
Float64Multiply(
first_row=0,
last_row=delta,
col=sid - 1,
first_col=sid - 1,
last_col=sid - 1,
value=1.0 / ratio,
)
)
zipline/data/_adjustments.pyx
+4 -4
View File
@@ -225,7 +225,7 @@ cpdef load_adjustments_from_sqlite(object adjustments_db, # sqlite3.Connection
asset_ixs[sid] = assets.get_loc(sid)
asset_ix = asset_ixs[sid]
price_adj = Float64Multiply(0, date_loc, asset_ix, ratio)
price_adj = Float64Multiply(0, date_loc, asset_ix, asset_ix, ratio)
for i, column in enumerate(columns):
col_adjustments = results[i]
if column != 'volume':
@@ -235,7 +235,7 @@ cpdef load_adjustments_from_sqlite(object adjustments_db, # sqlite3.Connection
col_adjustments[date_loc] = [price_adj]
else:
volume_adj = Float64Multiply(
0, date_loc, asset_ix, 1.0 / ratio
0, date_loc, asset_ix, asset_ix, 1.0 / ratio
)
try:
col_adjustments[date_loc].append(volume_adj)
@@ -253,7 +253,7 @@ cpdef load_adjustments_from_sqlite(object adjustments_db, # sqlite3.Connection
asset_ixs[sid] = assets.get_loc(sid)
asset_ix = asset_ixs[sid]
adj = Float64Multiply(0, date_loc, asset_ix, ratio)
adj = Float64Multiply(0, date_loc, asset_ix, asset_ix, ratio)
for i, column in enumerate(columns):
col_adjustments = results[i]
if column != 'volume':
@@ -273,7 +273,7 @@ cpdef load_adjustments_from_sqlite(object adjustments_db, # sqlite3.Connection
asset_ixs[sid] = assets.get_loc(sid)
asset_ix = asset_ixs[sid]
adj = Float64Multiply(0, date_loc, asset_ix, ratio)
adj = Float64Multiply(0, date_loc, asset_ix, asset_ix, ratio)
for i, column in enumerate(columns):
col_adjustments = results[i]
if column != 'volume':
zipline/lib/adjustment.pyx
+67 -33
View File
@@ -87,7 +87,7 @@ cpdef _from_assets_and_dates(cls,
end_date,
asset_id,
)
return cls(first_row, last_row, col, value)
return cls(first_row, last_row, col, col, value)
cdef class Float64Adjustment:
@@ -95,30 +95,37 @@ cdef class Float64Adjustment:
Base class for adjustments that operate on Float64 buffers.
"""
cdef:
readonly Py_ssize_t col, first_row, last_row
readonly Py_ssize_t first_col, last_col, first_row, last_row
readonly float64_t value
def __cinit__(self,
Py_ssize_t first_row,
Py_ssize_t last_row,
Py_ssize_t col,
Py_ssize_t first_col,
Py_ssize_t last_col,
object value):
assert 0 <= first_row <= last_row
assert 0 <= first_col <= last_col
self.first_row = first_row
self.last_row = last_row
self.col = col
self.first_col = first_col
self.last_col = last_col
self.value = float(value)
from_assets_and_dates = classmethod(_from_assets_and_dates)
def __repr__(self):
return "%s(first_row=%d, last_row=%d, col=%d, value=%f)" % (
type(self).__name__,
self.first_row,
self.last_row,
self.col,
self.value,
return (
"%s(first_row=%d, last_row=%d,"
" first_col=%d, last_col=%d, value=%f)" % (
type(self).__name__,
self.first_row,
self.last_row,
self.first_col,
self.last_col,
self.value,
)
)
def __richcmp__(self, object other, int op):
@@ -128,9 +135,15 @@ cdef class Float64Adjustment:
if op != Py_EQ or type(self) != type(other):
return NotImplemented
return self._key() == other._key()
cpdef _key(self):
return (
(self.first_row, self.last_row, self.col, self.value) == \
(other.first_row, other.last_row, other.col, other.value)
self.first_row,
self.last_row,
self.first_col,
self.last_col,
self.value,
)
@@ -148,21 +161,28 @@ cdef class Float64Multiply(Float64Adjustment):
[ 3., 4., 5.],
[ 6., 7., 8.]])
>>> adj = Float64Multiply(first_row=1, last_row=2, col=1, value=4.0)
>>> adj = Float64Multiply(
... first_row=1,
... last_row=2,
... first_col=1,
... last_col=2,
... value=4.0,
... )
>>> adj.mutate(arr)
>>> arr
array([[ 0., 1., 2.],
[ 3., 16., 5.],
[ 6., 28., 8.]])
[ 3., 16., 20.],
[ 6., 28., 32.]])
"""
cpdef mutate(self, float64_t[:, :] data):
cdef Py_ssize_t row, col
col = self.col
# last_row + 1 because last_row should also be affected.
for row in range(self.first_row, self.last_row + 1):
data[row, col] *= self.value
# last_col + 1 because last_col should also be affected.
for col in range(self.first_col, self.last_col + 1):
# last_row + 1 because last_row should also be affected.
for row in range(self.first_row, self.last_row + 1):
data[row, col] *= self.value
cdef class Float64Overwrite(Float64Adjustment):
@@ -179,21 +199,28 @@ cdef class Float64Overwrite(Float64Adjustment):
[ 3., 4., 5.],
[ 6., 7., 8.]])
>>> adj = Float64Overwrite(first_row=1, last_row=2, col=1, value=0.0)
>>> adj = Float64Overwrite(
... first_row=1,
... last_row=2,
... first_col=1,
... last_col=2,
... value=0.0,
... )
>>> adj.mutate(arr)
>>> arr
array([[ 0., 1., 2.],
[ 3., 0., 5.],
[ 6., 0., 8.]])
[ 3., 0., 0.],
[ 6., 0., 0.]])
"""
cpdef mutate(self, float64_t[:, :] data):
cdef Py_ssize_t row, col
col = self.col
# last_row + 1 because last_row should also be affected.
for row in range(self.first_row, self.last_row + 1):
data[row, col] = self.value
# last_col + 1 because last_col should also be affected.
for col in range(self.first_col, self.last_col + 1):
# last_row + 1 because last_row should also be affected.
for row in range(self.first_row, self.last_row + 1):
data[row, col] = self.value
cdef class Float64Add(Float64Adjustment):
@@ -210,18 +237,25 @@ cdef class Float64Add(Float64Adjustment):
[ 3., 4., 5.],
[ 6., 7., 8.]])
>>> adj = Float64Add(first_row=1, last_row=2, col=1, value=1.0)
>>> adj = Float64Add(
... first_row=1,
... last_row=2,
... first_col=1,
... last_col=2,
... value=1.0,
... )
>>> adj.mutate(arr)
>>> arr
array([[ 0., 1., 2.],
[ 3., 5., 5.],
[ 6., 8., 8.]])
[ 3., 5., 6.],
[ 6., 8., 9.]])
"""
cpdef mutate(self, float64_t[:, :] data):
cdef Py_ssize_t row, col
col = self.col
# last_row + 1 because last_row should also be affected.
for row in range(self.first_row, self.last_row + 1):
data[row, col] += self.value
# last_col + 1 because last_col should also be affected.
for col in range(self.first_col, self.last_col + 1):
# last_row + 1 because last_row should also be affected.
for row in range(self.first_row, self.last_row + 1):
data[row, col] += self.value