28fdecc98b
ENH: Make .latest return a Filter on bool columns.
2016-02-12 21:20:18 -05:00
d5bd2a9fb8
BLD: Pinning version of indirect dependency
2016-02-12 17:00:45 -05:00
fca7936b9b
Merge pull request #969 from quantopian/fix-numpy-warnings
...
Fix numpy warnings
2016-02-12 16:41:50 -05:00
24a292150a
MAINT: Removing confusing parameter where we only use default
2016-02-11 18:46:44 -05:00
86558f83b3
MAINT: use constants for nat
2016-02-11 18:46:43 -05:00
5eb453675d
BUG: don't fail if you cannot make a webrequest
2016-02-11 18:46:43 -05:00
69cc5e6564
MAINT: Delegate to math_utils to leave bottleneck as optional
2016-02-11 18:46:43 -05:00
318ac4deb5
TST: Use np.floor to preserve float return, since py3 returns an int
2016-02-11 18:46:43 -05:00
1075e0573e
PERF: Updated to sort_values for new pandas
2016-02-11 18:46:43 -05:00
069fe5ccda
TST: Updated for new pandas and made clearer types of values
2016-02-11 18:46:43 -05:00
f14002a741
TST: Test fixes for pandas update
2016-02-11 18:46:43 -05:00
ff485950b8
TST: Fix expected value now that assertion is more strict
...
in newer pandas
2016-02-11 18:46:43 -05:00
41231c5c59
MAINT: Removing unused functions and related tests
2016-02-11 18:46:43 -05:00
c11524fda1
MAINT: pandas 0.17 compat for blaze core loader
2016-02-11 18:46:43 -05:00
7f27df4c51
TST: We want the longest times, no matter the color
2016-02-11 18:46:42 -05:00
fda8daa120
PERF: Use iloc instead of values, which is slower on pandas 0.17
2016-02-11 18:46:42 -05:00
e33cc25265
TST: fix redundant partial
2016-02-11 18:46:42 -05:00
23c9bb0a7f
BLD: install system packages
2016-02-11 18:46:42 -05:00
68c9575b5c
BLD: remove conda stuff from travis
2016-02-11 18:46:42 -05:00
d94d9ae9d2
BLD: sed requirements file
2016-02-11 18:46:42 -05:00
d1eb479a19
BLD: add numpy 1.10 to build matrix
2016-02-11 18:46:42 -05:00
9a253d52b2
BLD: use numpy get_include
2016-02-11 18:46:42 -05:00
8be903b074
STY: remove unused import
2016-02-11 18:46:42 -05:00
83cc14db4c
BLD: run flake8 after tests
2016-02-11 18:46:42 -05:00
5a7bad4005
MAINT: Different workaround for old-style class.
2016-02-11 18:46:41 -05:00
5f49fa22cb
MAINT: Upgrade numpy and fix warnings.
...
Mostly fixes ambiguous calls to numpy.full, and uses explicitly-united
NaT values.
2016-02-11 18:46:39 -05:00
b65199339e
Merge pull request #991 from quantopian/fix-right-binding-filter-ops
...
BUG: Add right-binding operators to NumExprFilter.
2016-02-10 20:16:01 -05:00
9551e13895
DOC: Add whatsnew entry.
2016-02-10 18:51:33 -05:00
dfb1832231
TEST: More thorough testing of boolean binops.
2016-02-10 18:45:00 -05:00
e7269a880c
BUG: Add right-binding operators to NumExprFilter.
...
Fixes a bug where doing a boolean comparison between a non-numexpr
Filter and a NumExprFilter would fail because we don't implement
`__rand__` and `__ror__`.
2016-02-10 18:34:59 -05:00
9e89b4ab9b
Merge pull request #983 from quantopian/expiration_perf
...
Expiration perf
2016-02-05 10:02:06 -05:00
4e84b2c5ca
TST: Added doctest
2016-02-04 21:58:57 -05:00
79aff84aed
MAINT: Moved auto_close_date to Asset
...
Handle custom data source with non-int sids
2016-02-04 21:58:57 -05:00
ede1eb7aa0
PERF: Look up expired futures from in-memory Futures
...
instead of queries to the db.
2016-02-04 18:55:34 -05:00
f5b74bf7c5
Merge pull request #980 from quantopian/odo-kwargs
...
ENH: allow users to specify odo kwargs in from_blaze
2016-02-03 17:28:22 -05:00
7c8a44ecd7
ENH: allow users to specify odo kwargs in from_blaze
2016-02-03 16:16:06 -05:00
7173e9a5c8
Merge pull request #979 from quantopian/allow-cols
...
ENH: Allow passing of numpy arrays to writer.
2016-02-02 14:58:06 -05:00
27f94f83fa
ENH: Allow passing of numpy arrays to writer.
...
For faster parsing and writing workflows, do not require a DataFrame.
2016-02-02 14:03:42 -05:00
d2365750fd
Merge pull request #977 from quantopian/asset_comparison_overflow
...
BUG: Handle OverflowError by comparing False
2016-02-01 15:48:24 -05:00
9c51daad47
Merge pull request #978 from quantopian/allow-empty-writes
...
ENH: Add padding method to minute bars writer.
2016-02-01 15:22:10 -05:00
28b34f6409
BUG: Handle OverflowError by comparing False
...
instead of raising
2016-02-01 15:02:15 -05:00
488721e805
ENH: Add padding method to minute bars writer.
...
So that consumers can write empty days worth of data, without needing
to construct a DataFrame with zero data force a write.
The internal loader uses `last_date_in_output_for_sid` to signify that
data has been attempted to be retrieved for all dates up until that, so
that when resuming a job those retrieval of data for those dates are not
re-attempted.
Also, used to make the write logic cleaneer, by making it only
necessary to create an array large enough for the given df.
2016-02-01 14:18:22 -05:00
b102742f69
BUG: fix deltas in blaze core loader
2016-01-28 14:08:11 -05:00
dd64a6cb73
Merge pull request #974 from jlowin/patch-4
...
Replace print with logger.info
2016-01-25 19:43:19 -05:00
e44c0d42e1
Replace print with logger.info
2016-01-25 19:22:28 -05:00
6557553c39
Merge pull request #973 from quantopian/use-method-for-metadata
...
MAINT: Use metadata method for reader init.
2016-01-25 13:10:26 -05:00
930aa1b29b
MAINT: Use metadata method for reader init.
...
Use the preexisting metadata method when instantiating the minute bar
reader.
An internal sublcass uses the `_get_metadata` method to setup data for
directories that have not used the new writer/reader interface.
(i.e. allows for reader creation when the metadata.json file does not
exist.)
2016-01-25 12:58:43 -05:00
2803802736
Merge pull request #941 from quantopian/futures-tick-size
...
ENH: Adds tick_size and renames futures multiplier
2016-01-25 11:18:31 -05:00
e576231ae5
Merge pull request #972 from quantopian/fix-minute-bar-siblings
...
BUG: Fix OSError when creating sids that share dir
2016-01-25 10:41:42 -05:00
984e934e83
BUG: Fix OSError when creating sids that share dir
...
Fix a bug where creating a sid bcolz file when the containing directory
was already occupied by a sid caused an OSError on attempt of creating
the directory because it already existed.
e.g. if there were two sids, `1` and `2`. The paths would be
`00/00/000001.bcolz` and `00/00/000002.bcolz` which share the same
directory `00/00`.
Fixed by checking for directory existence before calling `makedirs`.
Add test coverage which exercises writing of sids that are siblings in
the sid directory structure.
2016-01-25 10:37:50 -05:00
Scott Sanderson