Python, JavaScript, C# and Fortran code for hosting EPA web applications and data/model services. Consult the wiki for details: https://github.com/quanted/qed/wiki Served publicly at:
These release notes cover the :ref:new features <whats-new-2.0>, as well as
some :ref:backwards incompatible changes <backwards-incompatible-2.0> you'll
want to be aware of when upgrading from Django 1.11 or earlier. We've
:ref:dropped some features<removed-features-2.0> that have reached the end of
their deprecation cycle, and we've :ref:begun the deprecation process for some features <deprecated-features-2.0>.
This release starts Django's use of a :ref:loose form of semantic versioning <internal-release-cadence>, but there aren't any major backwards incompatible
changes that might be expected of a 2.0 release. Upgrading should be a similar
amount of effort as past feature releases.
See the :doc:/howto/upgrade-version guide if you're updating an existing
project.
Python compatibility
Django 2.0 supports Python 3.4, 3.5, and 3.6. We highly recommend and only
officially support the latest release of each series.
The Django 1.11.x series is the last to support Python 2.7.
Django 2.0 will be the last release series to support Python 3.4. If you plan
a deployment of Python 3.4 beyond the end-of-life for Django 2.0 (April 2019),
stick with Django 1.11 LTS (supported until April 2020) instead. Note, however,
that the end-of-life for Python 3.4 is March 2019.
Third-party library support for older version of Django
Following the release of Django 2.0, we suggest that third-party app authors
drop support for all versions of Django prior to 1.11. At that time, you should
be able to run your package's tests using python -Wd so that deprecation
warnings do appear. After making the deprecation warning fixes, your app should
be compatible with Django 2.0.
.. _whats-new-2.0:
What's new in Django 2.0
Simplified URL routing syntax
The new :func:django.urls.path() function allows a simpler, more readable URL
routing syntax. For example, this example from previous Django releases::
The new syntax supports type coercion of URL parameters. In the example, the
view will receive the year keyword argument as an integer rather than as
a string. Also, the URLs that will match are slightly less constrained in the
rewritten example. For example, the year 10000 will now match since the year
integers aren't constrained to be exactly four digits long as they are in the
regular expression.
The django.conf.urls.url() function from previous versions is now available
as :func:django.urls.re_path. The old location remains for backwards
compatibility, without an imminent deprecation. The old
django.conf.urls.include() function is now importable from django.urls
so you can use from django.urls import include, path, re_path in your
URLconfs.
The :doc:/topics/http/urls document is rewritten to feature the new syntax
and provide more details.
Mobile-friendly contrib.admin
The admin is now responsive and supports all major mobile devices. Older
browsers may experience varying levels of graceful degradation.
Window expressions
The new :class:~django.db.models.expressions.Window expression allows
adding an OVER clause to querysets. You can use :ref:window functions <window-functions> and :ref:aggregate functions <aggregation-functions> in
the expression.
Minor features
:mod:django.contrib.admin
* The new :attr:`.ModelAdmin.autocomplete_fields` attribute and
:meth:`.ModelAdmin.get_autocomplete_fields` method allow using an
`Select2 <https://select2.org>`_ search widget for ``ForeignKey`` and
``ManyToManyField``.
:mod:`django.contrib.auth`
The default iteration count for the PBKDF2 password hasher is increased from
36,000 to 100,000.
:mod:django.contrib.gis
* Added MySQL support for the
:class:`~django.contrib.gis.db.models.functions.AsGeoJSON` function,
:class:`~django.contrib.gis.db.models.functions.GeoHash` function,
:class:`~django.contrib.gis.db.models.functions.IsValid` function,
:lookup:`isvalid` lookup, and :ref:`distance lookups <distance-lookups>`.
* Added the :class:`~django.contrib.gis.db.models.functions.Azimuth` and
:class:`~django.contrib.gis.db.models.functions.LineLocatePoint` functions,
supported on PostGIS and SpatiaLite.
* Any :class:`~django.contrib.gis.geos.GEOSGeometry` imported from GeoJSON now
has its SRID set.
* Added the :attr:`.OSMWidget.default_zoom` attribute to customize the map's
default zoom level.
* Made metadata readable and editable on rasters through the
:attr:`~django.contrib.gis.gdal.GDALRaster.metadata`,
:attr:`~django.contrib.gis.gdal.GDALRaster.info`, and
:attr:`~django.contrib.gis.gdal.GDALBand.metadata` attributes.
* Allowed passing driver-specific creation options to
:class:`~django.contrib.gis.gdal.GDALRaster` objects using ``papsz_options``.
* Allowed creating :class:`~django.contrib.gis.gdal.GDALRaster` objects in
GDAL's internal virtual filesystem. Rasters can now be :ref:`created from and
converted to binary data <gdal-raster-vsimem>` in-memory.
* The new :meth:`GDALBand.color_interp()
<django.contrib.gis.gdal.GDALBand.color_interp>` method returns the color
interpretation for the band.
:mod:`django.contrib.postgres`
The new distinct argument for
:class:~django.contrib.postgres.aggregates.ArrayAgg determines if
concatenated values will be distinct.
The new :class:~django.contrib.postgres.functions.RandomUUID database
function returns a version 4 UUID. It requires use of PostgreSQL's
pgcrypto extension which can be activated using the new
:class:~django.contrib.postgres.operations.CryptoExtension migration
operation.
:class:django.contrib.postgres.indexes.GinIndex now supports the
fastupdate and gin_pending_list_limit parameters.
The new :class:~django.contrib.postgres.indexes.GistIndex class allows
creating GiST indexes in the database. The new
:class:~django.contrib.postgres.operations.BtreeGistExtension migration
operation installs the btree_gist extension to add support for operator
classes that aren't built-in.
:djadmin:inspectdb can now introspect JSONField and various
RangeField\s (django.contrib.postgres must be in INSTALLED_APPS).
:mod:django.contrib.sitemaps
* Added the ``protocol`` keyword argument to the
:class:`~django.contrib.sitemaps.GenericSitemap` constructor.
Cache
cache.set_many() now returns a list of keys that failed to be inserted.
For the built-in backends, failed inserts can only happen on memcached.
File Storage
* :meth:`File.open() <django.core.files.File.open>` can be used as a context
manager, e.g. ``with file.open() as f:``.
Forms
The new date_attrs and time_attrs arguments for
:class:~django.forms.SplitDateTimeWidget and
:class:~django.forms.SplitHiddenDateTimeWidget allow specifying different
HTML attributes for the DateInput and TimeInput (or hidden)
subwidgets.
The new :meth:Form.errors.get_json_data() <django.forms.Form.errors.get_json_data> method returns form errors as
a dictionary suitable for including in a JSON response.
Generic Views
* The new :attr:`.ContextMixin.extra_context` attribute allows adding context
in ``View.as_view()``.
Management Commands
:djadmin:inspectdb now translates MySQL's unsigned integer columns to
PositiveIntegerField or PositiveSmallIntegerField.
The new :option:makemessages --add-location option controls the comment
format in PO files.
:djadmin:loaddata can now :ref:read from stdin <loading-fixtures-stdin>.
The new :option:diffsettings --output option allows formatting the output
in a unified diff format.
On Oracle, :djadmin:inspectdb can now introspect AutoField if the
column is created as an identity column.
On MySQL, :djadmin:dbshell now supports client-side TLS certificates.
Migrations
* The new :option:`squashmigrations --squashed-name` option allows naming the
squashed migration.
Models
The new :class:~django.db.models.functions.StrIndex database function
finds the starting index of a string inside another string.
On Oracle, AutoField and BigAutoField are now created as identity columns_.
The new chunk_size parameter of :meth:.QuerySet.iterator controls the
number of rows fetched by the Python database client when streaming results
from the database. For databases that don't support server-side cursors, it
controls the number of results Django fetches from the database adapter.
:meth:.QuerySet.earliest, :meth:.QuerySet.latest, and
:attr:Meta.get_latest_by <django.db.models.Options.get_latest_by> now
allow ordering by several fields.
Added the :class:~django.db.models.functions.ExtractQuarter function to
extract the quarter from :class:~django.db.models.DateField and
:class:~django.db.models.DateTimeField, and exposed it through the
:lookup:quarter lookup.
Added the :class:~django.db.models.functions.TruncQuarter function to
truncate :class:~django.db.models.DateField and
:class:~django.db.models.DateTimeField to the first day of a quarter.
Added the :attr:~django.db.models.Index.db_tablespace parameter to
class-based indexes.
If the database supports a native duration field (Oracle and PostgreSQL),
:class:~django.db.models.functions.Extract now works with
:class:~django.db.models.DurationField.
Added the of argument to :meth:.QuerySet.select_for_update(), supported
on PostgreSQL and Oracle, to lock only rows from specific tables rather than
all selected tables. It may be helpful particularly when
:meth:~.QuerySet.select_for_update() is used in conjunction with
:meth:~.QuerySet.select_related().
The new field_name parameter of :meth:.QuerySet.in_bulk allows fetching
results based on any unique model field.
:meth:.CursorWrapper.callproc() now takes an optional dictionary of keyword
parameters, if the backend supports this feature. Of Django's built-in
backends, only Oracle supports it.
The new :meth:connection.execute_wrapper() <django.db.backends.base.DatabaseWrapper.execute_wrapper> method allows
:doc:installing wrappers around execution of database queries </topics/db/instrumentation>.
The new filter argument for built-in aggregates allows :ref:adding different conditionals <conditional-aggregation> to multiple aggregations
over the same fields or relations.
Added support for expressions in :attr:Meta.ordering <django.db.models.Options.ordering>.
The new named parameter of :meth:.QuerySet.values_list allows fetching
results as named tuples.
The new :class:.FilteredRelation class allows adding an ON clause to
querysets.
Pagination
* Added :meth:`Paginator.get_page() <django.core.paginator.Paginator.get_page>`
to provide the documented pattern of handling invalid page numbers.
Requests and Responses
The :djadmin:runserver Web server supports HTTP 1.1.
Templates
* To increase the usefulness of :meth:`.Engine.get_default` in third-party
apps, it now returns the first engine if multiple ``DjangoTemplates`` engines
are configured in ``TEMPLATES`` rather than raising ``ImproperlyConfigured``.
* Custom template tags may now accept keyword-only arguments.
Tests
Added threading support to :class:~django.test.LiveServerTestCase.
Added settings that allow customizing the test tablespace parameters for
Oracle: :setting:DATAFILE_SIZE, :setting:DATAFILE_TMP_SIZE,
:setting:DATAFILE_EXTSIZE, and :setting:DATAFILE_TMP_EXTSIZE.
Validators
* The new :class:`.ProhibitNullCharactersValidator` disallows the null
character in the input of the :class:`~django.forms.CharField` form field
and its subclasses. Null character input was observed from vulnerability
scanning tools. Most databases silently discard null characters, but
psycopg2 2.7+ raises an exception when trying to save a null character to
a char/text field with PostgreSQL.
.. _backwards-incompatible-2.0:
Backwards incompatible changes in 2.0
=====================================
Removed support for bytestrings in some places
----------------------------------------------
To support native Python 2 strings, older Django versions had to accept both
bytestrings and unicode strings. Now that Python 2 support is dropped,
bytestrings should only be encountered around input/output boundaries (handling
of binary fields or HTTP streams, for example). You might have to update your
code to limit bytestring usage to a minimum, as Django no longer accepts
bytestrings in certain code paths.
For example, ``reverse()`` now uses ``str()`` instead of ``force_text()`` to
coerce the ``args`` and ``kwargs`` it receives, prior to their placement in
the URL. For bytestrings, this creates a string with an undesired ``b`` prefix
as well as additional quotes (``str(b'foo')`` is ``"b'foo'"``). To adapt, call
``decode()`` on the bytestring before passing it to ``reverse()``.
Database backend API
--------------------
This section describes changes that may be needed in third-party database
backends.
* The ``DatabaseOperations.datetime_cast_date_sql()``,
``datetime_cast_time_sql()``, ``datetime_trunc_sql()``,
``datetime_extract_sql()``, and ``date_interval_sql()`` methods now return
only the SQL to perform the operation instead of SQL and a list of
parameters.
* Third-party database backends should add a ``DatabaseWrapper.display_name``
attribute with the name of the database that your backend works with. Django
may use it in various messages, such as in system checks.
* The first argument of ``SchemaEditor._alter_column_type_sql()`` is now
``model`` rather than ``table``.
* The first argument of ``SchemaEditor._create_index_name()`` is now
``table_name`` rather than ``model``.
* To enable ``FOR UPDATE OF`` support, set
``DatabaseFeatures.has_select_for_update_of = True``. If the database
requires that the arguments to ``OF`` be columns rather than tables, set
``DatabaseFeatures.select_for_update_of_column = True``.
* To enable support for :class:`~django.db.models.expressions.Window`
expressions, set ``DatabaseFeatures.supports_over_clause`` to ``True``. You
may need to customize the ``DatabaseOperations.window_start_rows_start_end()``
and/or ``window_start_range_start_end()`` methods.
* Third-party database backends should add a
``DatabaseOperations.cast_char_field_without_max_length`` attribute with the
database data type that will be used in the
:class:`~django.db.models.functions.Cast` function for a ``CharField`` if the
``max_length`` argument isn't provided.
* The first argument of ``DatabaseCreation._clone_test_db()`` and
``get_test_db_clone_settings()`` is now ``suffix`` rather
than ``number`` (in case you want to rename the signatures in your backend
for consistency). ``django.test`` also now passes those values as strings
rather than as integers.
* Third-party database backends should add a
``DatabaseIntrospection.get_sequences()`` method based on the stub in
``BaseDatabaseIntrospection``.
Dropped support for Oracle 11.2
-------------------------------
The end of upstream support for Oracle 11.2 is Dec. 2020. Django 1.11 will be
supported until April 2020 which almost reaches this date. Django 2.0
officially supports Oracle 12.1+.
Default MySQL isolation level is read committed
-----------------------------------------------
MySQL's default isolation level, repeatable read, may cause data loss in
typical Django usage. To prevent that and for consistency with other databases,
the default isolation level is now read committed. You can use the
:setting:`DATABASES` setting to :ref:`use a different isolation level
<mysql-isolation-level>`, if needed.
:attr:`AbstractUser.last_name <django.contrib.auth.models.User.last_name>` ``max_length`` increased to 150
----------------------------------------------------------------------------------------------------------
A migration for :attr:`django.contrib.auth.models.User.last_name` is included.
If you have a custom user model inheriting from ``AbstractUser``, you'll need
to generate and apply a database migration for your user model.
If you want to preserve the 30 character limit for last names, use a custom
form::
from django.contrib.auth.forms import UserChangeForm
class MyUserChangeForm(UserChangeForm):
last_name = forms.CharField(max_length=30, required=False)
If you wish to keep this restriction in the admin when editing users, set
``UserAdmin.form`` to use this form::
from django.contrib.auth.admin import UserAdmin
from django.contrib.auth.models import User
class MyUserAdmin(UserAdmin):
form = MyUserChangeForm
admin.site.unregister(User)
admin.site.register(User, MyUserAdmin)
``QuerySet.reverse()`` and ``last()`` are prohibited after slicing
------------------------------------------------------------------
Calling ``QuerySet.reverse()`` or ``last()`` on a sliced queryset leads to
unexpected results due to the slice being applied after reordering. This is
now prohibited, e.g.::
>>> Model.objects.all()[:2].reverse()
Traceback (most recent call last):
...
TypeError: Cannot reverse a query once a slice has been taken.
Form fields no longer accept optional arguments as positional arguments
-----------------------------------------------------------------------
To help prevent runtime errors due to incorrect ordering of form field
arguments, optional arguments of built-in form fields are no longer accepted
as positional arguments. For example::
forms.IntegerField(25, 10)
raises an exception and should be replaced with::
forms.IntegerField(max_value=25, min_value=10)
``call_command()`` validates the options it receives
----------------------------------------------------
``call_command()`` now validates that the argument parser of the command being
called defines all of the options passed to ``call_command()``.
For custom management commands that use options not created using
``parser.add_argument()``, add a ``stealth_options`` attribute on the command::
class MyCommand(BaseCommand):
stealth_options = ('option_name', ...)
Indexes no longer accept positional arguments
---------------------------------------------
For example::
models.Index(['headline', '-pub_date'], 'index_name')
raises an exception and should be replaced with::
models.Index(fields=['headline', '-pub_date'], name='index_name')
Foreign key constraints are now enabled on SQLite
-------------------------------------------------
This will appear as a backwards-incompatible change (``IntegrityError:
FOREIGN KEY constraint failed``) if attempting to save an existing model
instance that's violating a foreign key constraint.
Foreign keys are now created with ``DEFERRABLE INITIALLY DEFERRED`` instead of
``DEFERRABLE IMMEDIATE``. Thus, tables may need to be rebuilt to recreate
foreign keys with the new definition, particularly if you're using a pattern
like this::
from django.db import transaction
with transaction.atomic():
Book.objects.create(author_id=1)
Author.objects.create(id=1)
If you don't recreate the foreign key as ``DEFERRED``, the first ``create()``
would fail now that foreign key constraints are enforced.
Backup your database first! After upgrading to Django 2.0, you can then
rebuild tables using a script similar to this::
from django.apps import apps
from django.db import connection
for app in apps.get_app_configs():
for model in app.get_models(include_auto_created=True):
if model._meta.managed and not (model._meta.proxy or model._meta.swapped):
for base in model.__bases__:
if hasattr(base, '_meta'):
base._meta.local_many_to_many = []
model._meta.local_many_to_many = []
with connection.schema_editor() as editor:
editor._remake_table(model)
This script hasn't received extensive testing and needs adaption for various
cases such as multiple databases. Feel free to contribute improvements.
In addition, because of a table alteration limitation of SQLite, it's prohibited
to perform :class:`~django.db.migrations.operations.RenameModel` and
:class:`~django.db.migrations.operations.RenameField` operations on models or
fields referenced by other models in a transaction. In order to allow migrations
containing these operations to be applied, you must set the
``Migration.atomic`` attribute to ``False``.
Miscellaneous
-------------
* The ``SessionAuthenticationMiddleware`` class is removed. It provided no
functionality since session authentication is unconditionally enabled in
Django 1.10.
* The default HTTP error handlers (``handler404``, etc.) are now callables
instead of dotted Python path strings. Django favors callable references
since they provide better performance and debugging experience.
* :class:`~django.views.generic.base.RedirectView` no longer silences
``NoReverseMatch`` if the ``pattern_name`` doesn't exist.
* When :setting:`USE_L10N` is off, :class:`~django.forms.FloatField` and
:class:`~django.forms.DecimalField` now respect :setting:`DECIMAL_SEPARATOR`
and :setting:`THOUSAND_SEPARATOR` during validation. For example, with the
settings::
USE_L10N = False
USE_THOUSAND_SEPARATOR = True
DECIMAL_SEPARATOR = ','
THOUSAND_SEPARATOR = '.'
an input of ``"1.345"`` is now converted to ``1345`` instead of ``1.345``.
* Subclasses of :class:`~django.contrib.auth.models.AbstractBaseUser` are no
longer required to implement ``get_short_name()`` and ``get_full_name()``.
(The base implementations that raise ``NotImplementedError`` are removed.)
``django.contrib.admin`` uses these methods if implemented but doesn't
require them. Third-party apps that use these methods may want to adopt a
similar approach.
* The ``FIRST_DAY_OF_WEEK`` and ``NUMBER_GROUPING`` format settings are now
kept as integers in JavaScript and JSON i18n view outputs.
* :meth:`~django.test.TransactionTestCase.assertNumQueries` now ignores
connection configuration queries. Previously, if a test opened a new database
connection, those queries could be included as part of the
``assertNumQueries()`` count.
* The default size of the Oracle test tablespace is increased from 20M to 50M
and the default autoextend size is increased from 10M to 25M.
* To improve performance when streaming large result sets from the database,
:meth:`.QuerySet.iterator` now fetches 2000 rows at a time instead of 100.
The old behavior can be restored using the ``chunk_size`` parameter. For
example::
Book.objects.iterator(chunk_size=100)
* Providing unknown package names in the ``packages`` argument of the
:class:`~django.views.i18n.JavaScriptCatalog` view now raises ``ValueError``
instead of passing silently.
* A model instance's primary key now appears in the default ``Model.__str__()``
method, e.g. ``Question object (1)``.
* ``makemigrations`` now detects changes to the model field ``limit_choices_to``
option. Add this to your existing migrations or accept an auto-generated
migration for fields that use it.
* Performing queries that require :ref:`automatic spatial transformations
<automatic-spatial-transformations>` now raises ``NotImplementedError``
on MySQL instead of silently using non-transformed geometries.
* ``django.core.exceptions.DjangoRuntimeWarning`` is removed. It was only used
in the cache backend as an intermediate class in ``CacheKeyWarning``'s
inheritance of ``RuntimeWarning``.
* Renamed ``BaseExpression._output_field`` to ``output_field``. You may need
to update custom expressions.
* In older versions, forms and formsets combine their ``Media`` with widget
``Media`` by concatenating the two. The combining now tries to :ref:`preserve
the relative order of elements in each list <form-media-asset-order>`.
``MediaOrderConflictWarning`` is issued if the order can't be preserved.
* ``django.contrib.gis.gdal.OGRException`` is removed. It's been an alias for
``GDALException`` since Django 1.8.
* Support for GEOS 3.3.x is dropped.
* The way data is selected for ``GeometryField`` is changed to improve
performance, and in raw SQL queries, those fields must now be wrapped in
``connection.ops.select``. See the :ref:`Raw queries note<gis-raw-sql>` in
the GIS tutorial for an example.
.. _deprecated-features-2.0:
Features deprecated in 2.0
==========================
``context`` argument of ``Field.from_db_value()`` and ``Expression.convert_value()``
------------------------------------------------------------------------------------
The ``context`` argument of ``Field.from_db_value()`` and
``Expression.convert_value()`` is unused as it's always an empty dictionary.
The signature of both methods is now::
(self, value, expression, connection)
instead of::
(self, value, expression, connection, context)
Support for the old signature in custom fields and expressions remains until
Django 3.0.
Miscellaneous
-------------
* The ``django.db.backends.postgresql_psycopg2`` module is deprecated in favor
of ``django.db.backends.postgresql``. It's been an alias since Django 1.9.
This only affects code that imports from the module directly. The
``DATABASES`` setting can still use
``'django.db.backends.postgresql_psycopg2'``, though you can simplify that by
using the ``'django.db.backends.postgresql'`` name added in Django 1.9.
* ``django.shortcuts.render_to_response()`` is deprecated in favor of
:func:`django.shortcuts.render`. ``render()`` takes the same arguments
except that it also requires a ``request``.
* The ``DEFAULT_CONTENT_TYPE`` setting is deprecated. It doesn't interact well
well with third-party apps and is obsolete since HTML5 has mostly superseded
XHTML.
* ``HttpRequest.xreadlines()`` is deprecated in favor of iterating over the
request.
* The ``field_name`` keyword argument to :meth:`.QuerySet.earliest` and
:meth:`.QuerySet.latest` is deprecated in favor of passing the field
names as arguments. Write ``.earliest('pub_date')`` instead of
``.earliest(field_name='pub_date')``.
.. _removed-features-2.0:
Features removed in 2.0
=======================
These features have reached the end of their deprecation cycle and are removed
in Django 2.0.
See :ref:`deprecated-features-1.9` for details on these changes, including how
to remove usage of these features.
* The ``weak`` argument to ``django.dispatch.signals.Signal.disconnect()`` is
removed.
* ``django.db.backends.base.BaseDatabaseOperations.check_aggregate_support()``
is removed.
* The ``django.forms.extras`` package is removed.
* The ``assignment_tag`` helper is removed.
* The ``host`` argument to ``SimpleTestCase.assertsRedirects()`` is removed.
The compatibility layer which allows absolute URLs to be considered equal to
relative ones when the path is identical is also removed.
* ``Field.rel`` and ``Field.remote_field.to`` are removed.
* The ``on_delete`` argument for ``ForeignKey`` and ``OneToOneField`` is now
required in models and migrations. Consider squashing migrations so that you
have less of them to update.
* ``django.db.models.fields.add_lazy_relation()`` is removed.
* When time zone support is enabled, database backends that don't support time
zones no longer convert aware datetimes to naive values in UTC anymore when
such values are passed as parameters to SQL queries executed outside of the
ORM, e.g. with ``cursor.execute()``.
* ``django.contrib.auth.tests.utils.skipIfCustomUser()`` is removed.
* The ``GeoManager`` and ``GeoQuerySet`` classes are removed.
* The ``django.contrib.gis.geoip`` module is removed.
* The ``supports_recursion`` check for template loaders is removed from:
* ``django.template.engine.Engine.find_template()``
* ``django.template.loader_tags.ExtendsNode.find_template()``
* ``django.template.loaders.base.Loader.supports_recursion()``
* ``django.template.loaders.cached.Loader.supports_recursion()``
* The ``load_template`` and ``load_template_sources`` template loader methods
are removed.
* The ``template_dirs`` argument for template loaders is removed:
* ``django.template.loaders.base.Loader.get_template()``
* ``django.template.loaders.cached.Loader.cache_key()``
* ``django.template.loaders.cached.Loader.get_template()``
* ``django.template.loaders.cached.Loader.get_template_sources()``
* ``django.template.loaders.filesystem.Loader.get_template_sources()``
* ``django.template.loaders.base.Loader.__call__()`` is removed.
* Support for custom error views that don't accept an ``exception`` parameter
is removed.
* The ``mime_type`` attribute of ``django.utils.feedgenerator.Atom1Feed`` and
``django.utils.feedgenerator.RssFeed`` is removed.
* The ``app_name`` argument to ``include()`` is removed.
* Support for passing a 3-tuple (including ``admin.site.urls``) as the first
argument to ``include()`` is removed.
* Support for setting a URL instance namespace without an application namespace
is removed.
* ``Field._get_val_from_obj()`` is removed.
* ``django.template.loaders.eggs.Loader`` is removed.
* The ``current_app`` parameter to the ``contrib.auth`` function-based views is
removed.
* The ``callable_obj`` keyword argument to
``SimpleTestCase.assertRaisesMessage()`` is removed.
* Support for the ``allow_tags`` attribute on ``ModelAdmin`` methods is
removed.
* The ``enclosure`` keyword argument to ``SyndicationFeed.add_item()`` is
removed.
* The ``django.template.loader.LoaderOrigin`` and
``django.template.base.StringOrigin`` aliases for
``django.template.base.Origin`` are removed.
See :ref:`deprecated-features-1.10` for details on these changes.
* The ``makemigrations --exit`` option is removed.
* Support for direct assignment to a reverse foreign key or many-to-many
relation is removed.
* The ``get_srid()`` and ``set_srid()`` methods of
``django.contrib.gis.geos.GEOSGeometry`` are removed.
* The ``get_x()``, ``set_x()``, ``get_y()``, ``set_y()``, ``get_z()``, and
``set_z()`` methods of ``django.contrib.gis.geos.Point`` are removed.
* The ``get_coords()`` and ``set_coords()`` methods of
``django.contrib.gis.geos.Point`` are removed.
* The ``cascaded_union`` property of ``django.contrib.gis.geos.MultiPolygon``
is removed.
* ``django.utils.functional.allow_lazy()`` is removed.
* The ``shell --plain`` option is removed.
* The ``django.core.urlresolvers`` module is removed in favor of its new
location, ``django.urls``.
* ``CommaSeparatedIntegerField`` is removed, except for support in historical
migrations.
* The template ``Context.has_key()`` method is removed.
* Support for the ``django.core.files.storage.Storage.accessed_time()``,
``created_time()``, and ``modified_time()`` methods is removed.
* Support for query lookups using the model name when
``Meta.default_related_name`` is set is removed.
* The MySQL ``__search`` lookup is removed.
* The shim for supporting custom related manager classes without a
``_apply_rel_filters()`` method is removed.
* Using ``User.is_authenticated()`` and ``User.is_anonymous()`` as methods
rather than properties is no longer supported.
* The ``Model._meta.virtual_fields`` attribute is removed.
* The keyword arguments ``virtual_only`` in ``Field.contribute_to_class()`` and
``virtual`` in ``Model._meta.add_field()`` are removed.
* The ``javascript_catalog()`` and ``json_catalog()`` views are removed.
* ``django.contrib.gis.utils.precision_wkt()`` is removed.
* In multi-table inheritance, implicit promotion of a ``OneToOneField`` to a
``parent_link`` is removed.
* Support for ``Widget._format_value()`` is removed.
* ``FileField`` methods ``get_directory_name()`` and ``get_filename()`` are
removed.
* The ``mark_for_escaping()`` function and the classes it uses: ``EscapeData``,
``EscapeBytes``, ``EscapeText``, ``EscapeString``, and ``EscapeUnicode`` are
removed.
* The ``escape`` filter now uses ``django.utils.html.conditional_escape()``.
* ``Manager.use_for_related_fields`` is removed.
* Model ``Manager`` inheritance follows MRO inheritance rules. The requirement
to use ``Meta.manager_inheritance_from_future`` to opt-in to the behavior is
removed.
* Support for old-style middleware using ``settings.MIDDLEWARE_CLASSES`` is
removed.
===========================
### 1.11.8
===========================
*December 2, 2017*
Django 1.11.8 fixes several bugs in 1.11.7.
Bugfixes
========
* Reallowed, following a regression in Django 1.10, ``AuthenticationForm`` to
raise the inactive user error when using ``ModelBackend`` (:ticket:`28645`).
* Added support for ``QuerySet.values()`` and ``values_list()`` for
``union()``, ``difference()``, and ``intersection()`` queries
(:ticket:`28781`).
* Fixed incorrect index name truncation when using a namespaced ``db_table``
(:ticket:`28792`).
* Made ``QuerySet.iterator()`` use server-side cursors on PostgreSQL after
``values()`` and ``values_list()`` (:ticket:`28817`).
* Fixed crash on SQLite and MySQL when ordering by a filtered subquery that
uses ``nulls_first`` or ``nulls_last`` (:ticket:`28848`).
* Made query lookups for ``CICharField``, ``CIEmailField``, and ``CITextField``
use a ``citext`` cast (:ticket:`28702`).
* Fixed a regression in caching of a ``GenericForeignKey`` when the referenced
model instance uses multi-table inheritance (:ticket:`28856`).
* Fixed "Cannot change column 'x': used in a foreign key constraint" crash on
MySQL with a sequence of ``AlterField`` and/or ``RenameField`` operations in
a migration (:ticket:`28305`).
===========================
html5lib 0.999999999 -> 1.0.1
1.0.1
Released on December 7, 2017
Breaking changes:
* Drop support for Python 2.6. (330) (Thank you, Hugo, Will Kahn-Greene!)
* Remove ``utils/spider.py`` (353) (Thank you, Jon Dufresne!)
Features:
* Improve documentation. (300, 307) (Thank you, Jon Dufresne, Tom Most,
Will Kahn-Greene!)
* Add iframe seamless boolean attribute. (Thank you, Ritwik Gupta!)
* Add itemscope as a boolean attribute. (194) (Thank you, Jonathan Vanasco!)
* Support Python 3.6. (333) (Thank you, Jon Dufresne!)
* Add CI support for Windows using AppVeyor. (Thank you, John Vandenberg!)
* Improve testing and CI and add code coverage (323, 334), (Thank you, Jon
Dufresne, John Vandenberg, Geoffrey Sneddon, Will Kahn-Greene!)
* Semver-compliant version number.
Bug fixes:
* Add support for setuptools < 18.5 to support environment markers. (Thank you,
John Vandenberg!)
* Add explicit dependency for six >= 1.9. (Thank you, Eric Amorde!)
* Fix regexes to work with Python 3.7 regex adjustments. (318, 379) (Thank
you, Benedikt Morbach, Ville Skyttä, Mark Vasilkov!)
* Fix alphabeticalattributes filter namespace bug. (324) (Thank you, Will
Kahn-Greene!)
* Include license file in generated wheel package. (350) (Thank you, Jon
Dufresne!)
* Fix annotation-xml typo. (339) (Thank you, Will Kahn-Greene!)
* Allow uppercase hex chararcters in CSS colour check. (377) (Thank you,
Komal Dembla, Hugo!)
### 1.0
Released and unreleased on December 7, 2017. Badly packaged release.
1.0b3
Released on July 24, 2013
* Removed ``RecursiveTreeWalker`` from ``treewalkers._base``. Any
implementation using it should be moved to
``NonRecursiveTreeWalker``, as everything bundled with html5lib has
for years.
* Fix 67 so that ``BufferedStream`` to correctly returns a bytes
object, thereby fixing any case where html5lib is passed a
non-seekable RawIOBase-like object.
### 1.0b2
Released on June 27, 2013
Removed reordering of attributes within the serializer. There is now
an alphabetical_attributes option which preserves the previous
behaviour through a new filter. This allows attribute order to be
preserved through html5lib if the tree builder preserves order.
Removed dom2sax from DOM treebuilders. It has been replaced by
treeadapters.sax.to_sax which is generic and supports any
treewalker; it also resolves all known bugs with dom2sax.
Fix treewalker assertions on hitting bytes strings on
Python 2. Previous to 1.0b1, treewalkers coped with mixed
bytes/unicode data on Python 2; this reintroduces this prior
behaviour on Python 2. Behaviour is unchanged on Python 3.
1.0b1
Released on May 17, 2013
* Implementation updated to implement the `HTML specification
<http://www.whatwg.org/specs/web-apps/current-work/>`_ as of 5th May
2013 (`SVN <http://svn.whatwg.org/webapps/>`_ revision r7867).
* Python 3.2+ supported in a single codebase using the ``six`` library.
* Removed support for Python 2.5 and older.
* Removed the deprecated Beautiful Soup 3 treebuilder.
``beautifulsoup4`` can use ``html5lib`` as a parser instead. Note that
since it doesn't support namespaces, foreign content like SVG and
MathML is parsed incorrectly.
* Removed ``simpletree`` from the package. The default tree builder is
now ``etree`` (using the ``xml.etree.cElementTree`` implementation if
available, and ``xml.etree.ElementTree`` otherwise).
* Removed the ``XHTMLSerializer`` as it never actually guaranteed its
output was well-formed XML, and hence provided little of use.
* Removed default DOM treebuilder, so ``html5lib.treebuilders.dom`` is no
longer supported. ``html5lib.treebuilders.getTreeBuilder("dom")`` will
return the default DOM treebuilder, which uses ``xml.dom.minidom``.
* Optional heuristic character encoding detection now based on
``charade`` for Python 2.6 - 3.3 compatibility.
* Optional ``Genshi`` treewalker support fixed.
* Many bugfixes, including:
* 33: null in attribute value breaks XML AttValue;
* 4: nested, indirect descendant, <button> causes infinite loop;
* `Google Code 215
<http://code.google.com/p/html5lib/issues/detail?id=215>`_: Properly
detect seekable streams;
* `Google Code 206
<http://code.google.com/p/html5lib/issues/detail?id=206>`_: add
support for <video preload=...>, <audio preload=...>;
* `Google Code 205
<http://code.google.com/p/html5lib/issues/detail?id=205>`_: add
support for <video poster=...>;
* `Google Code 202
<http://code.google.com/p/html5lib/issues/detail?id=202>`_: Unicode
file breaks InputStream.
* Source code is now mostly PEP 8 compliant.
* Test harness has been improved and now depends on ``nose``.
* Documentation updated and moved to https://html5lib.readthedocs.io/.
numba 0.35.0 -> 0.36.2
0.36.2
This is a bugfix release that provides minor changes to address:
PR 2645: Avoid CPython bug with exec in older 2.7.x.
PR 2652: Add support for CUDA 9.
0.36.1
This release continues to add new features to the work undertaken in partnership
with Intel on ParallelAccelerator technology. Other changes of note include the
compilation chain being updated to use LLVM 5.0 and the production of conda
packages using conda-build 3 and the new compilers that ship with it.
NOTE: A version 0.36.0 was tagged for internal use but not released.
ParallelAccelerator:
NOTE: The ParallelAccelerator technology is under active development and should
be considered experimental.
New features relating to ParallelAccelerator, from work undertaken with Intel,
include the addition of the stencil decorator for ease of implementation of
stencil-like computations, support for general reductions, and slice and
range fusion for parallel slice/bit-array assignments. Documentation on both the
use and implementation of the above has been added. Further, a new debug
environment variable NUMBA_DEBUG_ARRAY_OPT_STATS is made available to give
information about which operators/calls are converted to parallel for-loops.
ParallelAccelerator features:
PR 2457: Stencil Computations in ParallelAccelerator
PR 2548: Slice and range fusion, parallelizing bitarray and slice assignment
PR 2516: Support general reductions in ParallelAccelerator
ParallelAccelerator fixes:
PR 2540: Fix bug 2537
PR 2566: Fix issue 2564.
PR 2599: Fix nested multi-dimensional parfor type inference issue
PR 2604: Fixes for stencil tests and cmath sin().
PR 2605: Fixes issue 2603.
Additional features of note:
This release of Numba (and llvmlite) is updated to use LLVM version 5.0 as the
compiler back end, the main change to Numba to support this was the addition of
a custom symbol tracker to avoid the calls to LLVM's ExecutionEngine that was
crashing when asking for non-existent symbol addresses. Further, the conda
packages for this release of Numba are built using conda build version 3 and the
new compilers/recipe grammar that are present in that release.
PR 2568: Update for LLVM 5
PR 2607: Fixes abort when getting address to "nrt_unresolved_abort"
PR 2615: Working towards conda build 3
Thanks to community feedback and bug reports, the following fixes were also
made.
Misc fixes/enhancements:
PR 2534: Add tuple support to np.take.
PR 2551: Rebranding fix
PR 2552: relative doc links
PR 2570: Fix issue 2561, handle missing successor on loop exit
PR 2588: Fix 2555. Disable libpython.so linking on linux
PR 2601: Update llvmlite version dependency.
PR 2608: Fix potential cache file collision
PR 2612: Fix NRT test failure due to increased overhead when running in coverage
PR 2619: Fix dubious pthread_cond_signal not in lock
PR 2622: Fix np.nanmedian for all NaN case.
PR 2633: Fix markdown in CONTRIBUTING.md
PR 2635: Make the dependency on compilers for AOT optional.
CUDA support fixes:
PR 2523: Fix invalid cuda context in memory transfer calls in another thread
PR 2575: Use CPU to initialize xoroshiro states for GPU RNG. Fixes 2573
PR 2581: Fix cuda gufunc mishandling of scalar arg as array and out argument
pandas 0.21.0 -> 0.22.0
0.22.0
This is a major release from 0.21.1 and includes a single, API-breaking change.
We recommend that all users upgrade to this version after carefully reading the
release note (singular!).
.. _whatsnew_0220.api_breaking:
Backwards incompatible API changes
Pandas 0.22.0 changes the handling of empty and all-*NA* sums and products. The
summary is that
* The sum of an empty or all-*NA* ``Series`` is now ``0``
* The product of an empty or all-*NA* ``Series`` is now ``1``
* We've added a ``min_count`` parameter to ``.sum()`` and ``.prod()`` controlling
the minimum number of valid values for the result to be valid. If fewer than
``min_count`` non-*NA* values are present, the result is *NA*. The default is
``0``. To return ``NaN``, the 0.21 behavior, use ``min_count=1``.
Some background: In pandas 0.21, we fixed a long-standing inconsistency
in the return value of all-*NA* series depending on whether or not bottleneck
was installed. See :ref:`whatsnew_0210.api_breaking.bottleneck`. At the same
time, we changed the sum and prod of an empty ``Series`` to also be ``NaN``.
Based on feedback, we've partially reverted those changes.
Arithmetic Operations
^^^^^^^^^^^^^^^^^^^^^
The default sum for empty or all-*NA* ``Series`` is now ``0``.
*pandas 0.21.x*
.. code-block:: ipython
In [1]: pd.Series([]).sum()
Out[1]: nan
In [2]: pd.Series([np.nan]).sum()
Out[2]: nan
*pandas 0.22.0*
.. ipython:: python
pd.Series([]).sum()
pd.Series([np.nan]).sum()
The default behavior is the same as pandas 0.20.3 with bottleneck installed. It
also matches the behavior of NumPy's ``np.nansum`` on empty and all-*NA* arrays.
To have the sum of an empty series return ``NaN`` (the default behavior of
pandas 0.20.3 without bottleneck, or pandas 0.21.x), use the ``min_count``
keyword.
.. ipython:: python
pd.Series([]).sum(min_count=1)
Thanks to the ``skipna`` parameter, the ``.sum`` on an all-*NA*
series is conceptually the same as the ``.sum`` of an empty one with
``skipna=True`` (the default).
.. ipython:: python
pd.Series([np.nan]).sum(min_count=1) skipna=True by default
The ``min_count`` parameter refers to the minimum number of *non-null* values
required for a non-NA sum or product.
:meth:`Series.prod` has been updated to behave the same as :meth:`Series.sum`,
returning ``1`` instead.
.. ipython:: python
pd.Series([]).prod()
pd.Series([np.nan]).prod()
pd.Series([]).prod(min_count=1)
These changes affect :meth:`DataFrame.sum` and :meth:`DataFrame.prod` as well.
Finally, a few less obvious places in pandas are affected by this change.
Grouping by a Categorical
^^^^^^^^^^^^^^^^^^^^^^^^^
Grouping by a ``Categorical`` and summing now returns ``0`` instead of
``NaN`` for categories with no observations. The product now returns ``1``
instead of ``NaN``.
*pandas 0.21.x*
.. code-block:: ipython
In [8]: grouper = pd.Categorical(['a', 'a'], categories=['a', 'b'])
In [9]: pd.Series([1, 2]).groupby(grouper).sum()
Out[9]:
a 3.0
b NaN
dtype: float64
*pandas 0.22*
.. ipython:: python
grouper = pd.Categorical(['a', 'a'], categories=['a', 'b'])
pd.Series([1, 2]).groupby(grouper).sum()
To restore the 0.21 behavior of returning ``NaN`` for unobserved groups,
use ``min_count>=1``.
.. ipython:: python
pd.Series([1, 2]).groupby(grouper).sum(min_count=1)
Resample
^^^^^^^^
The sum and product of all-*NA* bins has changed from ``NaN`` to ``0`` for
sum and ``1`` for product.
*pandas 0.21.x*
.. code-block:: ipython
In [11]: s = pd.Series([1, 1, np.nan, np.nan],
...: index=pd.date_range('2017', periods=4))
...: s
Out[11]:
2017-01-01 1.0
2017-01-02 1.0
2017-01-03 NaN
2017-01-04 NaN
Freq: D, dtype: float64
In [12]: s.resample('2d').sum()
Out[12]:
2017-01-01 2.0
2017-01-03 NaN
Freq: 2D, dtype: float64
*pandas 0.22.0*
.. ipython:: python
s = pd.Series([1, 1, np.nan, np.nan],
index=pd.date_range('2017', periods=4))
s.resample('2d').sum()
To restore the 0.21 behavior of returning ``NaN``, use ``min_count>=1``.
.. ipython:: python
s.resample('2d').sum(min_count=1)
In particular, upsampling and taking the sum or product is affected, as
upsampling introduces missing values even if the original series was
entirely valid.
*pandas 0.21.x*
.. code-block:: ipython
In [14]: idx = pd.DatetimeIndex(['2017-01-01', '2017-01-02'])
In [15]: pd.Series([1, 2], index=idx).resample('12H').sum()
Out[15]:
2017-01-01 00:00:00 1.0
2017-01-01 12:00:00 NaN
2017-01-02 00:00:00 2.0
Freq: 12H, dtype: float64
*pandas 0.22.0*
.. ipython:: python
idx = pd.DatetimeIndex(['2017-01-01', '2017-01-02'])
pd.Series([1, 2], index=idx).resample("12H").sum()
Once again, the ``min_count`` keyword is available to restore the 0.21 behavior.
.. ipython:: python
pd.Series([1, 2], index=idx).resample("12H").sum(min_count=1)
Rolling and Expanding
^^^^^^^^^^^^^^^^^^^^^
Rolling and expanding already have a ``min_periods`` keyword that behaves
similar to ``min_count``. The only case that changes is when doing a rolling
or expanding sum with ``min_periods=0``. Previously this returned ``NaN``,
when fewer than ``min_periods`` non-*NA* values were in the window. Now it
returns ``0``.
*pandas 0.21.1*
.. code-block:: ipython
In [17]: s = pd.Series([np.nan, np.nan])
In [18]: s.rolling(2, min_periods=0).sum()
Out[18]:
0 NaN
1 NaN
dtype: float64
*pandas 0.22.0*
.. ipython:: python
s = pd.Series([np.nan, np.nan])
s.rolling(2, min_periods=0).sum()
The default behavior of ``min_periods=None``, implying that ``min_periods``
equals the window size, is unchanged.
Compatibility
If you maintain a library that should work across pandas versions, it
may be easiest to exclude pandas 0.21 from your requirements. Otherwise, all your
sum() calls would need to check if the Series is empty before summing.
With setuptools, in your setup.py use::
install_requires=['pandas!=0.21.*', ...]
With conda, use
.. code-block:: yaml
requirements:
run:
pandas !=0.21.0,!=0.21.1
Note that the inconsistency in the return value for all-NA series is still
there for pandas 0.20.3 and earlier. Avoiding pandas 0.21 will only help with
the empty case.
.. _whatsnew_0211:
0.21.1
This is a minor bug-fix release in the 0.21.x series and includes some small regression fixes,
bug fixes and performance improvements.
We recommend that all users upgrade to this version.
Highlights include:
Temporarily restore matplotlib datetime plotting functionality. This should
resolve issues for users who implicitly relied on pandas to plot datetimes
with matplotlib. See :ref:here <whatsnew_0211.converters>.
Improvements to the Parquet IO functions introduced in 0.21.0. See
:ref:here <whatsnew_0211.enhancements.parquet>.
.. contents:: What's new in v0.21.1
:local:
:backlinks: none
Pandas implements some matplotlib converters for nicely formatting the axis
labels on plots with ``datetime`` or ``Period`` values. Prior to pandas 0.21.0,
these were implicitly registered with matplotlib, as a side effect of ``import
pandas``.
In pandas 0.21.0, we required users to explicitly register the
converter. This caused problems for some users who relied on those converters
being present for regular ``matplotlib.pyplot`` plotting methods, so we're
temporarily reverting that change; pandas 0.21.1 again registers the converters on
import, just like before 0.21.0.
We've added a new option to control the converters:
``pd.options.plotting.matplotlib.register_converters``. By default, they are
registered. Toggling this to ``False`` removes pandas' formatters and restore
any converters we overwrote when registering them (:issue:`18301`).
We're working with the matplotlib developers to make this easier. We're trying
to balance user convenience (automatically registering the converters) with
import performance and best practices (importing pandas shouldn't have the side
effect of overwriting any custom converters you've already set). In the future
we hope to have most of the datetime formatting functionality in matplotlib,
with just the pandas-specific converters in pandas. We'll then gracefully
deprecate the automatic registration of converters in favor of users explicitly
registering them when they want them.
.. _whatsnew_0211.enhancements:
New features
.. _whatsnew_0211.enhancements.parquet:
Improvements to the Parquet IO functionality
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
:func:DataFrame.to_parquet will now write non-default indexes when the
underlying engine supports it. The indexes will be preserved when reading
back in with :func:read_parquet (:issue:18581).
:func:read_parquet now allows to specify the columns to read from a parquet file (:issue:18154)
:func:read_parquet now allows to specify kwargs which are passed to the respective engine (:issue:18216)
.. _whatsnew_0211.enhancements.other:
Other Enhancements
^^^^^^^^^^^^^^^^^^
:meth:Timestamp.timestamp is now available in Python 2.7. (:issue:17329)
:class:Grouper and :class:TimeGrouper now have a friendly repr output (:issue:18203).
.. _whatsnew_0211.deprecations:
Deprecations
- ``pandas.tseries.register`` has been renamed to
:func:`pandas.plotting.register_matplotlib_converters`` (:issue:`18301`)
.. _whatsnew_0211.performance:
Performance Improvements
Improved performance of plotting large series/dataframes (:issue:18236).
.. _whatsnew_0211.bug_fixes:
Bug Fixes
Conversion
^^^^^^^^^^
- Bug in :class:`TimedeltaIndex` subtraction could incorrectly overflow when ``NaT`` is present (:issue:`17791`)
- Bug in :class:`DatetimeIndex` subtracting datetimelike from DatetimeIndex could fail to overflow (:issue:`18020`)
- Bug in :meth:`IntervalIndex.copy` when copying and ``IntervalIndex`` with non-default ``closed`` (:issue:`18339`)
- Bug in :func:`DataFrame.to_dict` where columns of datetime that are tz-aware were not converted to required arrays when used with ``orient='records'``, raising``TypeError` (:issue:`18372`)
- Bug in :class:`DateTimeIndex` and :meth:`date_range` where mismatching tz-aware ``start`` and ``end`` timezones would not raise an err if ``end.tzinfo`` is None (:issue:`18431`)
- Bug in :meth:`Series.fillna` which raised when passed a long integer on Python 2 (:issue:`18159`).
Indexing
^^^^^^^^
- Bug in a boolean comparison of a ``datetime.datetime`` and a ``datetime64[ns]`` dtype Series (:issue:`17965`)
- Bug where a ``MultiIndex`` with more than a million records was not raising ``AttributeError`` when trying to access a missing attribute (:issue:`18165`)
- Bug in :class:`IntervalIndex` constructor when a list of intervals is passed with non-default ``closed`` (:issue:`18334`)
- Bug in ``Index.putmask`` when an invalid mask passed (:issue:`18368`)
- Bug in masked assignment of a ``timedelta64[ns]`` dtype ``Series``, incorrectly coerced to float (:issue:`18493`)
I/O
^^^
- Bug in class:`~pandas.io.stata.StataReader` not converting date/time columns with display formatting addressed (:issue:`17990`). Previously columns with display formatting were normally left as ordinal numbers and not converted to datetime objects.
- Bug in :func:`read_csv` when reading a compressed UTF-16 encoded file (:issue:`18071`)
- Bug in :func:`read_csv` for handling null values in index columns when specifying ``na_filter=False`` (:issue:`5239`)
- Bug in :func:`read_csv` when reading numeric category fields with high cardinality (:issue:`18186`)
- Bug in :meth:`DataFrame.to_csv` when the table had ``MultiIndex`` columns, and a list of strings was passed in for ``header`` (:issue:`5539`)
- Bug in parsing integer datetime-like columns with specified format in ``read_sql`` (:issue:`17855`).
- Bug in :meth:`DataFrame.to_msgpack` when serializing data of the ``numpy.bool_`` datatype (:issue:`18390`)
- Bug in :func:`read_json` not decoding when reading line deliminted JSON from S3 (:issue:`17200`)
- Bug in :func:`pandas.io.json.json_normalize` to avoid modification of ``meta`` (:issue:`18610`)
- Bug in :func:`to_latex` where repeated multi-index values were not printed even though a higher level index differed from the previous row (:issue:`14484`)
- Bug when reading NaN-only categorical columns in :class:`HDFStore` (:issue:`18413`)
- Bug in :meth:`DataFrame.to_latex` with ``longtable=True`` where a latex multicolumn always spanned over three columns (:issue:`17959`)
Plotting
^^^^^^^^
- Bug in ``DataFrame.plot()`` and ``Series.plot()`` with :class:`DatetimeIndex` where a figure generated by them is not pickleable in Python 3 (:issue:`18439`)
Groupby/Resample/Rolling
^^^^^^^^^^^^^^^^^^^^^^^^
- Bug in ``DataFrame.resample(...).apply(...)`` when there is a callable that returns different columns (:issue:`15169`)
- Bug in ``DataFrame.resample(...)`` when there is a time change (DST) and resampling frequecy is 12h or higher (:issue:`15549`)
- Bug in ``pd.DataFrameGroupBy.count()`` when counting over a datetimelike column (:issue:`13393`)
- Bug in ``rolling.var`` where calculation is inaccurate with a zero-valued array (:issue:`18430`)
Reshaping
^^^^^^^^^
- Error message in ``pd.merge_asof()`` for key datatype mismatch now includes datatype of left and right key (:issue:`18068`)
- Bug in ``pd.concat`` when empty and non-empty DataFrames or Series are concatenated (:issue:`18178` :issue:`18187`)
- Bug in ``DataFrame.filter(...)`` when :class:`unicode` is passed as a condition in Python 2 (:issue:`13101`)
- Bug when merging empty DataFrames when ``np.seterr(divide='raise')`` is set (:issue:`17776`)
Numeric
^^^^^^^
- Bug in ``pd.Series.rolling.skew()`` and ``rolling.kurt()`` with all equal values has floating issue (:issue:`18044`)
Categorical
^^^^^^^^^^^
- Bug in :meth:`DataFrame.astype` where casting to 'category' on an empty ``DataFrame`` causes a segmentation fault (:issue:`18004`)
- Error messages in the testing module have been improved when items have different ``CategoricalDtype`` (:issue:`18069`)
- ``CategoricalIndex`` can now correctly take a ``pd.api.types.CategoricalDtype`` as its dtype (:issue:`18116`)
- Bug in ``Categorical.unique()`` returning read-only ``codes`` array when all categories were ``NaN`` (:issue:`18051`)
- Bug in ``DataFrame.groupby(axis=1)`` with a ``CategoricalIndex`` (:issue:`18432`)
String
^^^^^^
- :meth:`Series.str.split()` will now propagate ``NaN`` values across all expanded columns instead of ``None`` (:issue:`18450`)
.. _whatsnew_0181:
pylint 1.7.4 -> 1.8.1
1.8
=========================
Release date: 2017-12-15
Respect disable=... in config file when running with --py3k.
New warning shallow-copy-environ added
Shallow copy of os.environ doesn't work as people may expect. os.environ
is not a dict object but rather a proxy object, so any changes made
on copy may have unexpected effects on os.environ
Instead of copy.copy(os.environ) method os.environ.copy() should be
used.
pyup-bot
commented
6 years ago
Updates
Here's a list of all the updates bundled in this pull request. I've added some links to make it easier for you to find all the information you need.
Changelogs
codecov 2.0.9 -> 2.0.10
django-parsley 0.6 -> 0.7
django 1.11.7 -> 2.0
html5lib 0.999999999 -> 1.0.1
numba 0.35.0 -> 0.36.2
pandas 0.21.0 -> 0.22.0
pylint 1.7.4 -> 1.8.1