Development¶
The homepage of joblib with user documentation is located on:
https://pythonhosted.org/joblib/
Getting the latest code¶
To get the latest code using git, simply type:
git clone git://github.com/joblib/joblib.git
If you don’t have git installed, you can download a zip or tarball of the latest code: http://github.com/joblib/joblib/archives/master
Installing¶
As any Python packages, to install joblib, simply do:
python setup.py install
in the source code directory.
Joblib has no other mandatory dependency than Python (supported versions are 2.6+ and 3.3+). Numpy (at least version 1.3) is an optional dependency for array manipulation.
Workflow to contribute¶
To contribute to joblib, first create an account on github. Once this is done, fork the joblib repository to have you own repository, clone it using ‘git clone’ on the computers where you want to work. Make your changes in your clone, push them to your github account, test them on several computer, and when you are happy with them, send a pull request to the main repository.
Running the test suite¶
To run the test suite, you need nosetests and the coverage modules. Run the test suite using:
nosetests
from the root of the project.
Building the docs¶
To build the docs you need to have setuptools and sphinx (>=0.5) installed. Run the command:
python setup.py build_sphinx
The docs are built in the build/sphinx/html directory.
Making a source tarball¶
To create a source tarball, eg for packaging or distributing, run the following command:
python setup.py sdist
The tarball will be created in the dist directory. This command will compile the docs, and the resulting tarball can be installed with no extra dependencies than the Python standard library. You will need setuptool and sphinx.
Making a release and uploading it to PyPI¶
This command is only run by project manager, to make a release, and upload in to PyPI:
python setup.py sdist bdist_egg bdist_wheel register upload
Updating the changelog¶
Changes are listed in the CHANGES.rst file. They must be manually updated but, the following git command may be used to generate the lines:
git log --abbrev-commit --date=short --no-merges --sparse
Licensing¶
joblib is BSD-licenced (3 clause):
This software is OSI Certified Open Source Software. OSI Certified is a certification mark of the Open Source Initiative.
Copyright (c) 2009-2011, joblib developpers All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
- Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
- Neither the name of Gael Varoquaux. nor the names of other joblib contributors may be used to endorse or promote products derived from this software without specific prior written permission.
This software is provided by the copyright holders and contributors “as is” and any express or implied warranties, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose are disclaimed. In no event shall the copyright owner or contributors be liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to, procurement of substitute goods or services; loss of use, data, or profits; or business interruption) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this software, even if advised of the possibility of such damage.
Latest changes¶
Release 0.9.0¶
Olivier Grisel
Make joblib use the ‘forkserver’ start method by default under Python 3.4+ to avoid causing crash with 3rd party libraries (such as Apple vecLib / Accelerate or the GCC OpenMP runtime) that use an internal thread pool that is not not reinitialized when afork
system call happens.
Olivier Grisel
New context manager based API (with
block) to re-use the same pool of workers across consecutive parallel calls.
Vlad Niculae and Olivier Grisel
Automated batching of fast tasks into longer running jobs to hide multiprocessing dispatching overhead when possible.
Loïc Estève
FIX compressed pickles interoperability between Python 2 and Python 3. Compressed pickles can be written with one and read with the other.
Olivier Grisel
FIX make it possible to calljoblib.load(filename, mmap_mode='r')
on pickled objects that include a mix of arrays of both memmory memmapable dtypes and object dtype.
Release 0.8.4¶
2014-11-20 Olivier Grisel
OPTIM use the C-optimized pickler under Python 3
This makes it possible to efficiently process parallel jobs that deal with numerous Python objects such as large dictionaries.
Release 0.8.3¶
2014-08-19 Olivier Grisel
FIX disable memmapping for object arrays
2014-08-07 Lars Buitinck
MAINT NumPy 1.10-safe version comparisons
2014-07-11 Olivier Grisel
FIX #146: Heisen test failure caused by thread-unsafe Python lists
This fix uses a queue.Queue datastructure in the failing test. This datastructure is thread-safe thanks to an internal Lock. This Lock instance not picklable hence cause the picklability check of delayed to check fail.
When using the threading backend, picklability is no longer required, hence this PRs give the user the ability to disable it on a case by case basis.
Release 0.8.2¶
2014-06-30 Olivier Grisel
BUG: use mmap_mode=’r’ by default in Parallel and MemmapingPool
The former default of mmap_mode=’c’ (copy-on-write) caused problematic use of the paging file under Windows.
2014-06-27 Olivier Grisel
BUG: fix usage of the /dev/shm folder under Linux
Release 0.8.0¶
2014-05-14 Olivier Grisel
Fix a bug in exception reporting under Python 3
2014-05-10 Olivier Grisel
Fixed a potential segfault when passing non-contiguous memmap instances.
2014-04-22 Gael Varoquaux
ENH: Make memory robust to modification of source files while the interpreter is running. Should lead to less spurious cache flushes and recomputations.
2014-02-24 Philippe Gervais
NewMemory.call_and_shelve
API to handle memoized results by reference instead of by value.
Release 0.8.0a3¶
2014-01-10 Olivier Grisel & Gael Varoquaux
FIX #105: Race condition in task iterable consumption when pre_dispatch != ‘all’ that could cause crash with error messages “Pools seems closed” and “ValueError: generator already executing”.
2014-01-12 Olivier Grisel
FIX #72: joblib cannot persist “output_dir” keyword argument.
Release 0.8.0a2¶
2013-12-23 Olivier Grisel
ENH: set default value of Parallel’s max_nbytes to 100MB
Motivation: avoid introducing disk latency on medium sized parallel workload where memory usage is not an issue.
FIX: properly handle the JOBLIB_MULTIPROCESSING env variable
FIX: timeout test failures under windows
Release 0.8.0a¶
2013-12-19 Olivier Grisel
FIX: support the new Python 3.4 multiprocessing API
2013-12-05 Olivier Grisel
ENH: make Memory respect mmap_mode at first call too
ENH: add a threading based backend to Parallel
This is low overhead alternative backend to the default multiprocessing backend that is suitable when calling compiled extensions that release the GIL.
Author: Dan Stahlke <dan@stahlke.org> Date: 2013-11-08
FIX: use safe_repr to print arg vals in trace
This fixes a problem in which extremely long (and slow) stack traces would be produced when function parameters are large numpy arrays.
2013-09-10 Olivier Grisel
ENH: limit memory copy with Parallel by leveraging numpy.memmap when possible
Release 0.7.1¶
2013-07-25 Gael Varoquaux
MISC: capture meaningless argument (n_jobs=0) in Parallel
2013-07-09 Lars Buitinck
ENH Handles tuples, sets and Python 3’s dict_keys type the same as lists. in pre_dispatch
2013-05-23 Martin Luessi
ENH: fix function caching for IPython
Release 0.7.0¶
This release drops support for Python 2.5 in favor of support for Python 3.0
2013-02-13 Gael Varoquaux
BUG: fix nasty hash collisions
2012-11-19 Gael Varoquaux
ENH: Parallel: Turn of pre-dispatch for already expanded lists
Gael Varoquaux 2012-11-19
ENH: detect recursive sub-process spawning, as when people do not protect the __main__ in scripts under Windows, and raise a useful error.
Gael Varoquaux 2012-11-16
ENH: Full python 3 support
Release 0.6.5¶
2012-09-15 Yannick Schwartz
BUG: make sure that sets and dictionnaries give reproducible hashes
2012-07-18 Marek Rudnicki
BUG: make sure that object-dtype numpy array hash correctly
2012-07-12 GaelVaroquaux
BUG: Bad default n_jobs for Parallel
Release 0.6.4¶
2012-05-07 Vlad Niculae
ENH: controlled randomness in tests and doctest fix
2012-02-21 GaelVaroquaux
ENH: add verbosity in memory
2012-02-21 GaelVaroquaux
BUG: non-reproducible hashing: order of kwargs
The ordering of a dictionnary is random. As a result the function hashing was not reproducible. Pretty hard to test
Release 0.6.3¶
2012-02-14 GaelVaroquaux
BUG: fix joblib Memory pickling
2012-02-11 GaelVaroquaux
BUG: fix hasher with Python 3
2012-02-09 GaelVaroquaux
API: filter_args: *args, **kwargs -> args, kwargs
Release 0.6.1¶
Bugfix release because of a merge error in release 0.6.0
Release 0.6.0¶
Beta 3
2012-01-11 Gael Varoquaux
BUG: ensure compatibility with old numpy
DOC: update installation instructions
BUG: file semantic to work under Windows
2012-01-10 Yaroslav Halchenko
BUG: a fix toward 2.5 compatibility
Beta 2
2012-01-07 Gael Varoquaux
ENH: hash: bugware to be able to hash objects defined interactively in IPython
2012-01-07 Gael Varoquaux
ENH: Parallel: warn and not fail for nested loops
ENH: Parallel: n_jobs=-2 now uses all CPUs but one
2012-01-01 Juan Manuel Caicedo Carvajal and Gael Varoquaux
ENH: add verbosity levels in Parallel
Release 0.5.7¶
2011-12-28 Gael varoquaux
API: zipped -> compress
2011-12-26 Gael varoquaux
ENH: Add a zipped option to Memory
API: Memory no longer accepts save_npy
2011-12-22 Kenneth C. Arnold and Gael varoquaux
BUG: fix numpy_pickle for array subclasses
2011-12-21 Gael varoquaux
ENH: add zip-based pickling
2011-12-19 Fabian Pedregosa
Py3k: compatibility fixes. This makes run fine the tests test_disk and test_parallel
Release 0.5.6¶
2011-12-11 Lars Buitinck
ENH: Replace os.path.exists before makedirs with exception check New disk.mkdirp will fail with other errnos than EEXIST.
2011-12-10 Bala Subrahmanyam Varanasi
MISC: pep8 compliant
Release 0.5.4¶
2011-09-29 Jon Olav Vik
BUG: Make mangling path to filename work on Windows
2011-09-25 Olivier Grisel
FIX: doctest heisenfailure on execution time
2011-08-24 Ralf Gommers
STY: PEP8 cleanup.
Release 0.5.2¶
2011-06-25 Gael varoquaux
ENH: Add cpu_count
2011-06-06 Gael varoquaux
ENH: Make sure memory hash in a reproducible way
Release 0.5.1¶
2011-04-12 Gael varoquaux
TEST: Better testing of parallel and pre_dispatch
Yaroslav Halchenko 2011-04-12
DOC: quick pass over docs – trailing spaces/spelling
Yaroslav Halchenko 2011-04-11
ENH: JOBLIB_MULTIPROCESSING env var to disable multiprocessing from the environment
Alexandre Gramfort 2011-04-08
ENH : adding log message to know how long it takes to load from disk the cache
Release 0.5.0¶
2011-04-01 Gael varoquaux
BUG: pickling MemoizeFunc does not store timestamp
2011-03-31 Nicolas Pinto
TEST: expose hashing bug with cached method
2011-03-26...2011-03-27 Pietro Berkes
BUG: fix error management in rm_subdirs BUG: fix for race condition during tests in mem.clear()
Gael varoquaux 2011-03-22...2011-03-26
TEST: Improve test coverage and robustness
Gael varoquaux 2011-03-19
BUG: hashing functions with only *var **kwargs
Gael varoquaux 2011-02-01... 2011-03-22
BUG: Many fixes to capture interprocess race condition when mem.cache is used by several processes on the same cache.
Fabian Pedregosa 2011-02-28
First work on Py3K compatibility
Gael varoquaux 2011-02-27
ENH: pre_dispatch in parallel: lazy generation of jobs in parallel for to avoid drowning memory.
GaelVaroquaux 2011-02-24
ENH: Add the option of overloading the arguments of the mother ‘Memory’ object in the cache method that is doing the decoration.
Gael varoquaux 2010-11-21
ENH: Add a verbosity level for more verbosity
Release 0.4.6¶
Gael varoquaux 2010-11-15
ENH: Deal with interruption in parallel
Gael varoquaux 2010-11-13
BUG: Exceptions raised by Parallel when n_job=1 are no longer captured.
Gael varoquaux 2010-11-13
BUG: Capture wrong arguments properly (better error message)
Release 0.4.5¶
Pietro Berkes 2010-09-04
BUG: Fix Windows peculiarities with path separators and file names BUG: Fix more windows locking bugs
Gael varoquaux 2010-09-03
ENH: Make sure that exceptions raised in Parallel also inherit from the original exception class ENH: Add a shadow set of exceptions
Fabian Pedregosa 2010-09-01
ENH: Clean up the code for parallel. Thanks to Fabian Pedregosa for the patch.
Release 0.4.4¶
Gael varoquaux 2010-08-23
BUG: Fix Parallel on computers with only one CPU, for n_jobs=-1.
Gael varoquaux 2010-08-02
BUG: Fix setup.py for extra setuptools args.
Gael varoquaux 2010-07-29
MISC: Silence tests (and hopefuly Yaroslav :P)
Release 0.4.3¶
Gael Varoquaux 2010-07-22
BUG: Fix hashing for function with a side effect modifying their input argument. Thanks to Pietro Berkes for reporting the bug and proving the patch.
Release 0.4.2¶
Gael Varoquaux 2010-07-16
BUG: Make sure that joblib still works with Python2.5. => release 0.4.2