|
1
|
|
|
# Copyright 2015 Quantopian, Inc. |
|
2
|
|
|
# |
|
3
|
|
|
# Licensed under the Apache License, Version 2.0 (the "License"); |
|
4
|
|
|
# you may not use this file except in compliance with the License. |
|
5
|
|
|
# You may obtain a copy of the License at |
|
6
|
|
|
# |
|
7
|
|
|
# http://www.apache.org/licenses/LICENSE-2.0 |
|
8
|
|
|
# |
|
9
|
|
|
# Unless required by applicable law or agreed to in writing, software |
|
10
|
|
|
# distributed under the License is distributed on an "AS IS" BASIS, |
|
11
|
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
|
12
|
|
|
# See the License for the specific language governing permissions and |
|
13
|
|
|
# limitations under the License. |
|
14
|
|
|
from numpy import ( |
|
15
|
|
|
iinfo, |
|
16
|
|
|
uint32, |
|
17
|
|
|
) |
|
18
|
|
|
|
|
19
|
|
|
from zipline.data.us_equity_pricing import ( |
|
20
|
|
|
BcolzDailyBarReader, |
|
21
|
|
|
SQLiteAdjustmentReader, |
|
22
|
|
|
) |
|
23
|
|
|
from zipline.lib.adjusted_array import ( |
|
24
|
|
|
adjusted_array, |
|
25
|
|
|
) |
|
26
|
|
|
from zipline.errors import NoFurtherDataError |
|
27
|
|
|
|
|
28
|
|
|
from .base import PipelineLoader |
|
29
|
|
|
|
|
30
|
|
|
UINT32_MAX = iinfo(uint32).max |
|
31
|
|
|
|
|
32
|
|
|
|
|
33
|
|
|
class USEquityPricingLoader(PipelineLoader): |
|
34
|
|
|
""" |
|
35
|
|
|
PipelineLoader for US Equity Pricing data |
|
36
|
|
|
|
|
37
|
|
|
Delegates loading of baselines and adjustments. |
|
38
|
|
|
""" |
|
39
|
|
|
|
|
40
|
|
|
def __init__(self, raw_price_loader, adjustments_loader): |
|
41
|
|
|
self.raw_price_loader = raw_price_loader |
|
42
|
|
|
# HACK: Pull the calendar off our raw_price_loader so that we can |
|
43
|
|
|
# backshift dates. |
|
44
|
|
|
self._calendar = self.raw_price_loader._calendar |
|
45
|
|
|
self.adjustments_loader = adjustments_loader |
|
46
|
|
|
|
|
47
|
|
|
@classmethod |
|
48
|
|
|
def from_files(cls, pricing_path, adjustments_path): |
|
49
|
|
|
""" |
|
50
|
|
|
Create a loader from a bcolz equity pricing dir and a SQLite |
|
51
|
|
|
adjustments path. |
|
52
|
|
|
|
|
53
|
|
|
Parameters |
|
54
|
|
|
---------- |
|
55
|
|
|
pricing_path : str |
|
56
|
|
|
Path to a bcolz directory written by a BcolzDailyBarWriter. |
|
57
|
|
|
adjusments_path : str |
|
58
|
|
|
Path to an adjusments db written by a SQLiteAdjustmentWriter. |
|
59
|
|
|
""" |
|
60
|
|
|
return cls( |
|
61
|
|
|
BcolzDailyBarReader(pricing_path), |
|
62
|
|
|
SQLiteAdjustmentReader(adjustments_path) |
|
63
|
|
|
) |
|
64
|
|
|
|
|
65
|
|
|
def load_adjusted_array(self, columns, dates, assets, mask): |
|
66
|
|
|
# load_adjusted_array is called with dates on which the user's algo |
|
67
|
|
|
# will be shown data, which means we need to return the data that would |
|
68
|
|
|
# be known at the start of each date. We assume that the latest data |
|
69
|
|
|
# known on day N is the data from day (N - 1), so we shift all query |
|
70
|
|
|
# dates back by a day. |
|
71
|
|
|
start_date, end_date = _shift_dates( |
|
72
|
|
|
self._calendar, dates[0], dates[-1], shift=1, |
|
73
|
|
|
) |
|
74
|
|
|
|
|
75
|
|
|
raw_arrays = self.raw_price_loader.load_raw_arrays( |
|
76
|
|
|
columns, |
|
77
|
|
|
start_date, |
|
78
|
|
|
end_date, |
|
79
|
|
|
assets, |
|
80
|
|
|
) |
|
81
|
|
|
adjustments = self.adjustments_loader.load_adjustments( |
|
82
|
|
|
columns, |
|
83
|
|
|
dates, |
|
84
|
|
|
assets, |
|
85
|
|
|
) |
|
86
|
|
|
adjusted_arrays = [ |
|
87
|
|
|
adjusted_array(raw_array, mask, col_adjustments) |
|
88
|
|
|
for raw_array, col_adjustments in zip(raw_arrays, adjustments) |
|
89
|
|
|
] |
|
90
|
|
|
|
|
91
|
|
|
return dict(zip(columns, adjusted_arrays)) |
|
92
|
|
|
|
|
93
|
|
|
|
|
94
|
|
|
def _shift_dates(dates, start_date, end_date, shift): |
|
95
|
|
|
try: |
|
96
|
|
|
start = dates.get_loc(start_date) |
|
97
|
|
|
except KeyError: |
|
98
|
|
|
if start_date < dates[0]: |
|
99
|
|
|
raise NoFurtherDataError( |
|
100
|
|
|
msg=( |
|
101
|
|
|
"Pipeline Query requested data starting on {query_start}, " |
|
102
|
|
|
"but first known date is {calendar_start}" |
|
103
|
|
|
).format( |
|
104
|
|
|
query_start=str(start_date), |
|
105
|
|
|
calendar_start=str(dates[0]), |
|
106
|
|
|
) |
|
107
|
|
|
) |
|
108
|
|
|
else: |
|
109
|
|
|
raise ValueError("Query start %s not in calendar" % start_date) |
|
110
|
|
|
|
|
111
|
|
|
# Make sure that shifting doesn't push us out of the calendar. |
|
112
|
|
|
if start < shift: |
|
113
|
|
|
raise NoFurtherDataError( |
|
114
|
|
|
msg=( |
|
115
|
|
|
"Pipeline Query requested data from {shift}" |
|
116
|
|
|
" days before {query_start}, but first known date is only " |
|
117
|
|
|
"{start} days earlier." |
|
118
|
|
|
).format(shift=shift, query_start=start_date, start=start), |
|
119
|
|
|
) |
|
120
|
|
|
|
|
121
|
|
|
try: |
|
122
|
|
|
end = dates.get_loc(end_date) |
|
123
|
|
|
except KeyError: |
|
124
|
|
|
if end_date > dates[-1]: |
|
125
|
|
|
raise NoFurtherDataError( |
|
126
|
|
|
msg=( |
|
127
|
|
|
"Pipeline Query requesting data up to {query_end}, " |
|
128
|
|
|
"but last known date is {calendar_end}" |
|
129
|
|
|
).format( |
|
130
|
|
|
query_end=end_date, |
|
131
|
|
|
calendar_end=dates[-1], |
|
132
|
|
|
) |
|
133
|
|
|
) |
|
134
|
|
|
else: |
|
135
|
|
|
raise ValueError("Query end %s not in calendar" % end_date) |
|
136
|
|
|
return dates[start - shift], dates[end - shift] |
|
137
|
|
|
|
quantopian /
zipline
