(math/R-RNetCDF) Updated 2.4.2 to 2.5.2

Version 2.5-2, 2021-08-20
  * Support hdf5 filters via multi-filter interface (netcdf>=4.8.0)
  * Windows: update binary packages to netcdf 4.7.4 with OpenDAP
  * Generate type conversions with m4 macros
  * Reduce CPU time for utcal.nc example to pass CRAN checks

(mef)

doc: Updated math/R-pbkrtest to 0.5.1

(mef)

(math/R-pbkrtest)  Updated 0.4.7 to 0.5.1

pbkrtest v0.5.1 (Release date: 2021-03-09)
============================================

Changes

* Improved documentation

pbkrtest v0.5-0.0 (Release date: 2020-08-04)
============================================

Changes

* Satterthwaite approximation added via the SATmodcomp function.

* Checks for models being nested is not performed for parametric
  bootstrap any longer. Reason is that the simr package use parametric
  bootstrap for testing variance components being zero.

* doi added to DESCRIPTION file

pbkrtest v0.4-8.6 (Release date: 2020-02-20)
============================================

Bug fixes:

* documentation fixed ddf_Lb is now exported

* mclapply issue for windows fixed

* vcovAdj.lmerMod is exported to make emmeans work. Contact Russ Lenth
  to make emmeans used generic function vcovAdj.

pbkrtest v0.4-8 (Release date: 2020-02-20)
==========================================

Bug fixes:

  * Issue related to class() versus inherits() fixed.

Changes:

  * NEWS file added
  * NAMESPACE file is now generated automatically

(mef)

doc: Updated math/R-DEoptimR to 1.0.9

(mef)

(math/R-DEoptimR) Updated 1.0.8 to 1.0.9, NEWS not updated

(mef)

doc: remove devel/ruby-cucumber-tag-expressions to 4.0.2 line

(taca)

devel/ruby-cucumber-tag-expressions: revert to 3.0.1

Revert to 3.0.1 from previous update of 4.0.2 since ruby-cucumber-core
require ruby-cucumber-tag-expressions < 4.

(taca)

Reinstate the comment that the 5.6 limitation is for php-mysql, because that's
valuable insight, and still unfortunately the case.

No functional change.

(dsainty)

Added math/R-broom version 0.7.9

(mef)

(math/R-broom) import R-broom-0.7.9

Summarizes key information about statistical objects in tidy tibbles.
This makes it easy to report results, create plots and consistently
work with large numbers of models at once. Broom provides three verbs
that each provide different types of information about a model. tidy()
summarizes information about model components such as coefficients of
a regression. glance() reports information about an entire model, such
as goodness of fit measures like AIC and BIC. augment() adds
information about individual observations to a dataset, such as fitted
values or influence measures.

(mef)

Alpha is 64 bits.

(jklos)

doc/TODO: + gimp-2.10.28.

(wiz)

doc: Updated finance/ruby-activemerchant to 1.123.0

(taca)

finance/ruby-activemerchant: update to 1.122.0

Version 1.122.0 (2021-08-03)

* Orbital: Correct success logic for refund [tatsianaclifton] #4014
* usaepay: Added pin gateway setting [DustinHaefele] #4026
* MercadoPago: Added external_reference, more payer object options, and
  metadata field [DustinHaefele] #4020
* Element: Add duplicate_override_flag [almalee24] #4012
* PayTrace: Support gateway [meagabeth] #3985
* vPOS: Support credit + refund [therufs] #3998
* PayArc: Support gateway [senthil-code] #3974
* NMI: Support cardholder_auth field for 3DS2 [cdmackeyfree] #4002
* Confiable: Support cardtype [therufs] #4004
* Maestro: Add BIN [therufs] #4003
* PayULatam: Ensure phone number is pulled from shipping_address correctly
  [dsmcclain] #4005
* SafeCharge: Add challenge_preference for 3DS [klaiv] #3999
* Adyen: Pass networkTxReference in all transactions [naashton] #4006
* Adyen: Ensure correct transaction reference is selected [dsmcclain] #4007
* PayTrace: Support level_3_data fields [meagabeth] #4008
* BluePay: Add support for Stored Credentials [dsmcclain] #4009
* Orbital: Add support for SCARecurringPayment [jessiagee] #4010
* Braintree: Support recurring_first and moto reasons [curiousepic] #4013
* PayTrace: Adjust capture method [meagabeth] #4015
* BarclaysEpdqExtraPlus: updated custom_eci test + remote tests [yyapuncich]
  #4022
* CyberSource: Add customerID field [deemeyers] #4025
* CyberSource: Adjust Auth [naashton] #3956
* Valid Canadian Institution Numbers [naashton] #4024
* PayTrace: Adjust purchase and capture methods to handle MultiResponse
  scenarios [meagabeth] #4027
* Payflow: Add support for MERCHDESCR field [rachelkirk] #4028
* PayTrace: Support $0 authorize in verify method [meagabeth] #4030
* PayArc: Add error_code in response [cdm-83] #4021
* Update bank routing account validation check [jessiagee] #4029
* Kushki: Add 'contactDetails' fields [mbreenlyles] #4033
* Adyen: Truncating order_id and remote test [yyapuncich] #4036
* CyberSource: Allow string content for Ignore AVS/CVV flags [curiousepic]
  #4043
* Decidir: Update validation error message handling [arbianchi] #4042
* Authorize.net: Remove cardholderAuthentication for non-3DS transactions
  [BritneyS] #4045
* BlueSnap: Handle 429 errors [britth] #4044
* Orbital: Update unit test files [meagabeth] #4046
* Orbital: Strip null characters from responses [britth] #4041
* Merchant Warrior: Handle invalid XML responses [arbianchi] #4047
* Braintree: Fix NoMethodError for failed card verification [molbrown] #4048
* Worldpay: Accepting 3DS1 and 3DS2 authentication data from external MPI
  [chandan-PS] #4017
* PayArc: Currency and parameters updates [jessiagee] #4051
* Elavon: Add support for special characters [mbreenlyles] #4049
* PayArc: Formatting CC month, adding tax_rate, removing default void reason
  [jessiagee] #4053
* Kushki: Add support for fullResponse field [rachelkirk] #4057
* Element: Add support for MerchantDescriptor field [BritneyS] #4058
* PayArc: Added email and phone to credit and charge [jessiagee] #4056
* Mundipagg: Added support for 'authentication_secret_key' for 'api_key'
  overwrite [DustinHaefele] #4059
* Payflow: Raise an error if store method is called [dsmcclain] #4066
* Monei: JSON API implementation [jimmyn] #3613
* Maestro: Update BINs [therufs] #4067
* Monei: Change domain to monei.com [jimmyn] #4068
* Spreedly: Support gateway_specific_response_fields in response params
  [abarrak] #4064
* Payeezy: Add support for add_soft_descriptors [rachelkirk] #4069
* Stripe Payment Intents: Add support for network_transaction_id field
  [cdmackeyfree] #4060
* Worldpay: Support 'CAPTURED' response for authorize transactions
  [naashton] #4070
* Ingenico (Global Collect): New idempotence key header [BritneyS] #4073
* PayTrace: Adjust handling of line_items subfields [meagabeth] #4074
* Worldpay: Correct Expiration Year Format [tatsianaclifton] #4076
* Monei: Improve Scrub Regex [tatsianaclifton] #4072
* Payflow: add THREEDSVERSION and DSTRANSACTIONID when present [bbraschi]
  #4075

1.121.0 (2021-06-09)

* Braintree: Lift restriction on gem version to allow for backwards
  compatibility [naashton] #3993
* Payment Express/Windcave: Send amount on verify calls [cdmackeyfree] #3995
* Orbital: Use billing_address name as fallback [curiousepic] #3966
* vPOS: handle shop_process_id correctly [therufs] #3996
* Checkout v2: Support metadata field [saschakala] #3992
* Adyen: Support networkTxReference field [naashton] #3997
* Paypal Express: Enable PayPal express reference transaction request to
  send merchant session id [janees-e] #3994

(taca)

doc: Updated devel/ruby-test-unit to 3.4.7

(taca)

devel/ruby-test-unit: update to 3.4.7

3.4.7 - 2021-09-14

Fixes

* Suppressed a warning on Ruby 2.  [GitHub#205][Patch by Kenichi Kamiya]

Thanks

Kenichi Kamiya

3.4.6 - 2021-09-11

Improvements

* Added support for tests that use Ractor.  Use Test::Unit::TestCase.ractor
  to declare that these tests use Ractor.

* Added --debug-on-failure option.

3.4.5 - 2021-09-04

Improvements

* Added more metadata to gemspec.  [GitHub#183][Patch by Kenichi Kamiya]

* Removed needless files from gem.  [GitHub#184][Patch by Kenichi Kamiya]

* Updated documents.
  [GitHub#191][GitHub#192][GitHub#193][GitHub#199][GitHub#200]
  [GitHub#201][Patch by Kenichi Kamiya]

Added assert_nothing_leaked_memory.

Fixes

* Fixed typos in documents.
  [GitHub#189][GitHub#190][GitHub#195][GitHub#197][Patch by Kenichi Kamiya]

Thanks

* Kenichi Kamiya

(taca)

doc: Updated devel/ruby-gettext to 3.4.1

(taca)

devel/ruby-gettext: update to 3.4.1

3.4.1: 2021-09-08

Improvements

* Listed licenses separately in gemspec. [GitHub#88][Patch by Josh Cooper]

Thanks

* Josh Cooper

(taca)

mikutter: explicitly pull the latest ruby-gtk2 to avoid dependency woes.

Bump PKGREVISION.

(tsutsui)

doc: Updated devel/ruby-ffi to 1.15.4

(taca)

devel/ruby-ffi: update to 1.15.4

1.15.4 / 2021-09-01
Fixed:

* Fix build for uClibc. #913
* Correct module lookup when including ffi-module gem. #912

Changed:

* Use ruby code of the ffi gem in JRuby-9.2.20+. #915

(taca)

doc: Updated devel/ruby-cucumber-tag-expressions to 4.0.2

(taca)

devel/ruby-cucumber-tag-expressions: update to 4.0.2

4.0.0 (2021-09-02)

* use new berp for all platforms
* Use canned berp recipes for all platforms
* make grammar file location configurable
* Rename master to main #1507

4.0.2 (2021-09-13)

* No change except version

(taca)

ham/trustedQSL: Drop superceded patch.

The patch to add a NetBSD-specific accomoodation for PATH_MAX seems
not really right, as POSIX specifies limits.h.  It is not necessary
for building, even if one removes upstream's workaround.

To be pointed out upstream, and we can re-remediate if necessary.

NFCI

(gdt)

doc: Updated graphics/R-colourpicker to 1.1.0

(mef)

(graphics/R-colourpicker) Updated 1.0 to 1.1.0

# Colourpicker 1.1.0 (2020-08-19)

- `colourInput()` now works with all R colour names, including names
  that have aliases such as "gray" vs "grey" (#25)
- add support for bookmarking (restoring state) (#33)
- added "closeOnClick" parameter that, when `TRUE`, will cause the
  colour selection panel to close immediately after choosing a colour
  (#34)
- Fix issue where transparent colours no longer showed the background
  as transparent (#40)
- Fix the colour picker addin and Plot Colour Helper addin that were
  broken due to web security issues
- Use {shinydisconnect} for demo app

(mef)

ham/trustedQSL: reduce patches

This commit does not intend to make any functional changes.

  - drop removal of LIB_SUFFIX via SUBST, as that seems overtaken by
    upstream

  - set db5 include prefix correctly and drop patches to work around
    the old way

(gdt)

doc: Updated graphics/R-farver to 2.1.0

(mef)

(graphics/R-farver) Updated 2.0.3 to 2.1.0

# farver 2.1.0

* Added support for OKLab and OKLch (polar version of OKLab) colour spaces
* Added `encode_native()` and `decode_native()` to convert back and forth
  between the internal R colour format.

(mef)

doc: Updated devel/R-waldo to 0.3.1

(mef)

(devel/waldo) Updated 0.3.0 to 0.3 1

# waldo 0.3.1

* `compare()`ing data frames now works independently of `option(max.print)`
  (#105).

* Fixed regression when comparing vectors with missing values (#102).

(mef)

doc: Updated devel/R-xfun to 0.26

(mef)

(math/R-xfun) Updated 0.25 to 0.26

# CHANGES IN xfun VERSION 0.26

- The `windows_only` argument of `native_encode()` has been
  removed. Now `native_encode()` only tries the conversion to native
  encoding on platforms where `l10n_info()[['UTF-8']]` does not return
  `TRUE`.

- Added a `solaris` argument to `upload_win_builder()`.

(mef)

doc: Updated misc/R-pillar to 1.6.2

(mef)

(misc/R-pillar) Updated 1.6.1 to 1.6.2

# pillar 1.6.2
## Options

- Options `pillar.print_max`, `pillar.print_min`, `pillar.width` and
  `pillar.max_extra_cols` are now queried before the corresponding
  `tibble.` or `dplyr.` options are consulted, the latter will be
  soft-deprecated in pillar v2.0.0 (#353).
- New `pillar.bidi` option. When active, control characters are
  inserted to improve display of data with right-to-left text (#333).
- The new `pillar.max_footer_lines` option (default: 7) allows
  controlling the maximum number of footer lines shown. It is applied
  in addition to the existing `tibble.max_extra_cols` option (#263).

## Formatting
- If a column doesn't make use of all horizontal width offered to it,
  the excess width is distributed over other columns (#331).
- Improved allocation of free space in multi-tier tables with
  `getOption("tibble.width") > getOption("width")` (#344).
- All pillars are shown with their true horizontal extent,
  irrespective of the indicated `width`. This simplifies the
  implementation of custom `pillar_shaft()` methods (#347).

## Features
- `num()` gains `extra_sigfig` argument to automatically show more
  significant figures for numbers of the same magnitude with subtle
  differences (#97).
- `print.tbl()` and `format.tbl()` support the `max_extra_cols` and
  `max_footer_lines` arguments that override the corresponding options
  (#360).
- `print.tbl()` and `format.tbl()` maps the now deprecated `n_extra`
  argument to `max_extra_cols` for consistency (#360).

## Bug fixes
- Avoid mangling of duplicate column names in footer (#332).
- Fix warning with zero of type `bit64::integer64()` (#319).

## Documentation
- All package options are now documented in `?pillar_options` (#339).
- `obj_sum()` no longer calls `type_sum()` for vectors since pillar
  v1.6.1, this is now documented (#321).
- Fix documentation on usage of `vctrs::vec_proxy()` and
  `vctrs::vec_restore()` (#322).

## Internal
- Using `attr(exact = TRUE)` everywhere.
- `is_vector_s3()` is no longer generic (#181).
- Fix internal logic around `vec_proxy()` and `vec_restore()` (#316).

(mef)

doc: Updated math/R-units to 0.7.2

(mef)

(math/units)  Updated 0.7.1 to 0.7.2

# version 0.7-2

* enhance `pillar` integration; #273, #275, #280 @krlmlr

* new `unique` method for `units` and `mixed_units` objects; #283 addressing
  #277 @lewinfox

(mef)

doc: Updated devel/R-BH to 1.75.0.0

(mef)

(devel/R-BH) Updated 1.69.0.1 to 1.75.0.0

2020-12-24  Dirk Eddelbuettel  <edd@debian.org>

* .github/workflows/ci.yaml: Small tweaks to CI YAML file

2020-12-12  Dirk Eddelbuettel  <edd@debian.org>

* DESCRIPTION (Version, Date): Release 1.75.0-0

* inst/include/boost/*: Upgraded to Boost 1.75

* README.md: Updated package list for 1.75
* DESCRIPTION (Description): Idem

* local/scripts/CreateBoost.sh: Update for 1.75.0, added beast

* .github/workflows/ci.yaml: Add CI runner using r-ci
* README.md: Add new badge

* README.md: Update several URLs to https
* inst/NEWS.Rd: Idem

2020-05-03  Dirk Eddelbuettel  <edd@debian.org>

* inst/LICENSE_1_0.txt: Included from upstream tar.gz

2020-03-31  Dirk Eddelbuettel  <edd@debian.org>

* README.md: Minor badge rearrangement, added example

2020-03-30  Dirk Eddelbuettel  <edd@debian.org>

* README.md: Added badges, edited, fixed typo

2020-03-15  Dirk Eddelbuettel  <edd@debian.org>

* DESCRIPTION (Version, Date): Roll minor version

* NAMESPACE: Removed as not needed without R code

2019-12-28  Dirk Eddelbuettel  <edd@debian.org>

* DESCRIPTION (Version, Date): Release 1.72.0-3

* local/scripts/CreateBoost.sh (boostinstall): Remove old boost
installation to always create a fresh installation

2019-12-16  Dirk Eddelbuettel  <edd@debian.org>

* DESCRIPTION (Version, Date): Release 1.72.0-2

* inst/include/boost/numeric/odeint/stepper/: Correct accidental
renaming in this directory, only rename in .../generation/

* DESCRIPTION (URL): Added a link to GH repo

2019-12-15  Dirk Eddelbuettel  <edd@debian.org>

* DESCRIPTION (Version, Date): Release 1.72.0-1

* inst/include/boost/*: Upgraded to Boost 1.72

* README.md: Updated package list for 1.72

* inst/include/boost/numeric/odeint/stepper/generation/: As before
rename karp54_classic.hpp to karp54_cl.hpp, adjust generation.hpp
* inst/include/boost/numeric/ublas/storage.hpp : Add missing include
for boost/serialization/array_wrapper.hpp

* inst/include/boost/numeric/odeint/stepper/generation/: As before
New adjust for generation_controlled_adams_bashforth_moulton.hpp
* inst/include/boost/math/special_functions/hypergeometric_1F1.hpp:
Adjust to renamed hypergeometric_1F1_small_a_negative_b_by_ratio.hpp

* inst/include/boost/*: As before disable all diagnostic pragmas in
accordance with the CRAN Policy (though this may not be a good idea...)

(mef)

doc: Updated math/R-tidyr to 1.1.3

(mef)

(math/R-tidyr) Updated 1.0.2 to 1.1.3

# tidyr 1.1.3

* tidyr verbs no longer have "default" methods for lazyeval fallbacks. This
  means that you'll get clearer error messages (#1036).

* `uncount()` error for non-integer weights and gives a clearer error message
  for negative weights (@mgirlich, #1069).

* You can once again unnest dates (#1021, #1089).

* `pivot_wider()` works with data.table and empty key variables (@mgirlich, #1066).

* `separate_rows()` works for factor columns (@mgirlich, #1058).

# tidyr 1.1.2

* `separate_rows()` returns to 1.1.0 behaviour for empty strings
  (@rjpatm, #1014).

# tidyr 1.1.1

* New tidyr logo!

* stringi dependency has been removed; this was a substantial dependency that
  make tidyr hard to compile in resource constrained environments
  (@rjpat, #936).

* Replace Rcpp with cpp11. See <https://cpp11.r-lib.org/articles/motivations.html>
  for reasons why.

# tidyr 1.1.0

## General features

* `pivot_longer()`, `hoist()`, `unnest_wider()`, and `unnest_longer()` gain
  new `transform` arguments; these allow you to transform values "in flight".
  They are partly needed because vctrs coercion rules have become stricter,
  but they give you greater flexibility than was available previously (#921).

* Arguments that use tidy selection syntax are now clearly documented and
  have been updated to use tidyselect 1.1.0 (#872).

## Pivoting improvements

* Both `pivot_wider()` and `pivot_longer()` are considerably more performant,
  thanks largely to improvements in the underlying vctrs code
  (#790, @DavisVaughan).

* `pivot_longer()` now supports `names_to = character()` which prevents the
  name column from being created (#961).

    ```{r}
    df <- tibble(id = 1:3, x_1 = 1:3, x_2 = 4:6)
    df %>% pivot_longer(-id, names_to = character())
    ```

* `pivot_longer()` no longer creates a `.copy` variable in the presence of
  duplicate column names. This makes it more consistent with the handling
  of non-unique specs.

* `pivot_longer()` automatically disambiguates non-unique ouputs, which can
  occur when the input variables include some additional component that you
  don't care about and want to discard (#792, #793).

    ```{r}
    df <- tibble(id = 1:3, x_1 = 1:3, x_2 = 4:6)
    df %>% pivot_longer(-id, names_pattern = "(.)_.")
    df %>% pivot_longer(-id, names_sep = "_", names_to = c("name", NA))
    df %>% pivot_longer(-id, names_sep = "_", names_to = c(".value", NA))
    ```

* `pivot_wider()` gains a `names_sort` argument which allows you to sort
  column names in order. The default, `FALSE`, orders columms by their
  first appearance (#839). In a future version, I'll consider changing the
  default to `TRUE`.

* `pivot_wider()` gains a `names_glue` argument that allows you to construct
  output column names with a glue specification.

* `pivot_wider()` arguments `values_fn` and `values_fill` can now be single
  values; you now only need to use a named list if you want to use different
  values for different value columns (#739, #746). They also get improved
  errors if they're not of the expected type.

## Rectangling

* `hoist()` now automatically names pluckers that are a single string (#837).
  It error if you use duplicated column names (@mgirlich, #834), and now uses
  `rlang::list2()` behind the scenes (which means that you can now use `!!!`
  and `:=`) (#801).

* `unnest_longer()`, `unnest_wider()`, and `hoist()` do a better job
  simplifying list-cols. They no longer add unneeded `unspecified()` when
  the result is still a list (#806), and work when the list contains
  non-vectors (#810, #848).

* `unnest_wider(names_sep = "")` now provides default names for unnamed inputs,
  suppressing the many previous name repair messages (#742).

## Nesting

* `pack()` and `nest()` gains a `.names_sep` argument allows you to strip outer
  names from inner names, in symmetrical way to how the same argument to
  `unpack()` and `unnest()` combines inner and outer names (#795, #797).

* `unnest_wider()` and `unnest_longer()` can now unnest `list_of` columns. This
  is important for unnesting columns created from `nest()` and with
  `pivot_wider()`, which will create `list_of` columns if the id columns are
  non-unique (#741).

## Bug fixes and minor improvements

* `chop()` now creates list-columns of class `vctrs::list_of()`. This helps
  keep track of the type in case the chopped data frame is empty, allowing
  `unchop()` to reconstitute a data frame with the correct number and types
  of column even when there are no observations.

* `drop_na()` now preserves attributes of unclassed vectors (#905).

* `expand()`, `expand_grid()`, `crossing()`, and `nesting()` once again
  evaluate their inputs iteratively, so you can refer to freshly created
  columns, e.g. `crossing(x = seq(-2, 2), y = x)` (#820).

* `expand()`, `expand_grid()`, `crossing()`, and `nesting()` gain a
  `.name_repair` giving you control over their name repair strategy
  (@jeffreypullin, #798).

* `extract()` lets you use `NA` in `into`, as documented (#793).

* `extract()`, `separate()`, `hoist()`, `unnest_longer()`, and `unnest_wider()`
  give a better error message if `col` is missing (#805).

* `pack()`'s first argument is now `.data` instead of `data` (#759).

* `pivot_longer()` now errors if `values_to` is not a length-1 character vector
  (#949).

* `pivot_longer()` and `pivot_wider()` are now generic so implementations
  can be provided for objects other than data frames (#800).

* `pivot_wider()` can now pivot data frame columns (#926)

* `unite(na.rm = TRUE)` now works for all types of variable, not just character
  vectors (#765).

* `unnest_wider()` gives a better error message if you attempt to unnest
  multiple columns (#740).

* `unnest_auto()` works when the input data contains a column called `col`
  (#959).

(mef)

doc: Updated devel/ruby-cucumber-messages to 17.1.1

(taca)

devel/ruby-cucumber-messages: update to 17.1.1

No change except version.

(taca)

Updated devel/p5-Proc-ProcessTable to 0.62

(wen)

doc: Updated devel/ruby-cucumber-expressions to 12.1.3

(taca)

devel/ruby-cucumber-expressions: update to 12.1.3

No change except version.

(taca)

Update to 0.62

Upstream changes:
0.62 2021-09-13
  - updated Changes file to stick to perl5-module-meta recommendations
  - changed version scheme back to 2 digits after major version.
  - added J. Bargsten as author

0.612 2021-09-08
  - regression issue: allow (again) processes with empty names on Linux

0.611 2021-08-18
  - fixed potential memory leak while parsing command names in linux

0.61 2021-08-17
  - fixed regression: on linux only 1st char of the command names was parsed

0.60 2021-08-14
  - compiles on musl
  - Skip LXC hidden dir in /dev when looking for TTYs
  - Fix test on OpenBSD

(wen)

doc: Updated databases/ruby-sequel to 5.48.0

(taca)

databases/ruby-sequel: update to 5.48.0

5.48.0 (2021-09-01)

* Make the unused_associations plugin association reflection tracking work
  correctly when combining coverage runs (jeremyevans)
* Add Database#like_without_collate on MSSQL, to avoid using COLLATE on LIKE
  arguments, which can significantly improve performance (jeremyevans)
* Add Model::Errors#full_message private method for easiest i18n support for
  errors with multiple attributes (jeremyevans) (#1779)

(taca)

Updated devel/p5-Proc-Background to 1.22

(wen)

Update to 1.22

Upstream changes:
Version 1.22

  * Fix bug in timeout of $p->wait on Unix perl earlier than 5.12
    where Time::HiRes::alarm was failing to set the alarm.
  * Fix test case 47 (which was failing on about 5% of systems)
    by waiting longer for the process to exit

(wen)

doc: Updated converters/ruby-unf_ext to 0.0.8

(taca)

converters/ruby-unf_ext: update to 0.0.8

0.0.8 (2021-09-14)

* No functional change in the library code.
* Include Windows binaries for Ruby 3.0.
* Drop support for Ruby 2.1 and earlier.
* Replace Travis CI with Github Actions.
* Fix cross-build after upgrading rake-compiler/rake-compiler-dock to
  1.1.1/1.1.0.

(taca)

doc/TODO: update sudo

+ sudo-1.9.8p1.

(taca)

doc: Updated archivers/ruby-ffi-libarchive to 1.1.3

(taca)

archivers/ruby-ffi-libarchive: update to 1.1.3

1.1.3 (2021-09-16)

* Update chefstyle requirement from 1.6.2 to 1.7.1 #52
  (dependabot-preview[bot])
* Update chefstyle requirement from 1.7.1 to 1.7.2 #53
  (dependabot-preview[bot])
* Update chefstyle requirement from 1.7.2 to 1.7.5 #55
  (dependabot-preview[bot])
* Upgrade to GitHub-native Dependabot #56 (dependabot-preview[bot])
* Remove support for EOL Ruby 2.4 #59 (tas50)
* Update chefstyle requirement from 1.7.5 to 2.0.8 #60 (dependabot[bot])
* Update chefstyle requirement from 2.0.8 to 2.0.9 #61 (dependabot[bot])
* Add strip components support #58 (jasonwbarnett)

(taca)

doc: Updated print/R-knitr to 1.34

(mef)

(print/R-knitr) Updated 1.33 to 1.34

(https://github.com/yihui/knitr/releases)
@yihui yihui released this Sep 9, 2021

NEW FEATURES

  * Added a package option latex.tilde so that users can customize the tilde
    symbol, e.g., you can set knitr::opts_knit$set(latex.tilde = "\\hlopt{\\
    scriptsize{\\ensuremath{\\sim}}}"). This is because the default tilde is
    too high in the PDF output, and it's often more desirable to use a middle
    tilde instead (thanks, @brry #1992, @jaredlander #492).

  * For the tikz engine, the class options of the standalone document classs
    can be specified via the chunk option engine.opts$classoption (thanks,
    @XiangyunHuang, #1985). The default value is tikz, i.e., \documentclass
    [tikz]{standalone} is used by default.

  * Added the ability to pass additional arguments to dvisvgm when using the
    tikz engine with fig.ext = "svg" by using dvisvgm.opts in engine.opts
    (thanks, @andrewheiss, #2039). Recent versions of dvisvgm now allow you to
    embed fonts into SVG files as base64-encoded WOFF files, so tikz chunks can
    embed fonts using like so: ```{tikz, fig.ext="svg", engine.opts=list
    (dvisvgm.opts = "--font-format=woff")}.

  * Added a new targets engine (ropensci/targets#503, @wlandau). Details:
    https://books.ropensci.org/targets/markdown.html.

  * The chunk option cache.globals can take values TRUE and FALSE now (in
    addition to a character vector). When FALSE, it tries to find all symbols
    in the code chunk, no matter if they are local or global variables. When
    NULL or TRUE, it tries to find all global variables (thanks, @knokknok, #
    1898).

MAJOR CHANGES

  * An error is now thrown when an inline code result is not coercible to
    character. This has always been the assumed behavior but it happens to be
    different in certain formats with unknown automatic coercion. This is now
    enforced to prevent any unexpected output. An inline code expression must
    evaluate to a character vector or an object coercible by as.character() (#
    2006).

  * The markdown package has been moved from Imports to Suggests in knitr's
    DESCRIPTION, which means it is no longer a hard dependency but has become a
    soft dependency of knitr (#1864). One consequence for package developers is
    that if you use the vignette engines based on markdown (such as
    knitr::knitr), you will have to explicitly declare the (soft) dependency on
    markdown, because the implicit dependency through knitr is no longer there.

  * For R packages that use the knitr::rmarkdown engine to build their
    vignettes, rmarkdown must be declared as a dependency in the package
    DESCRIPTION (e.g., in Suggests). Previously, knitr would fall back to using
    the markdown package to build vignettes if rmarkdown is not declared (#1864
    , #2020).

  * write_bib() only uses the first URL if multiple are found in a package (#
    2028).

MINOR CHANGES

  * The attribute data-external="1" will be added to <iframe>s generated by
    include_url() to prevent Pandoc from embedding the URL as base64 data
    (thanks, @IndrajeetPatil, https://stackoverflow.com/q/67477667/559676).

  * The chunk option strip.white = TRUE used to work only when the chunk option
    collapse = FALSE. Now the two options are independent, i.e., strip.white
    also works when collapse = TRUE (thanks, @kbvernon, #2011).

  * When building R Markdown vignettes but Pandoc is not available, the
    vignette engine will emit a message instead of a warning before falling
    back to using the markdown package.

  * The internal function is_abs_path() has been removed. Users (if any) should
    use the exported function xfun::is_abs_path() instead.

BUG FIXES

  * Fix an issue with the RStudio IDE when using knitr::include_url() or
    knitr::include_app() in interactive Notebook mode. This will no more cause
    an error but print the list object as is (thanks, @systemnova, #2015).

  * Fix a regression with fig.keep chunk option used in chunks with only one
    figure where the figure was not showing in output (thanks, @fmichonneau, #
    1993).

  * Allow vignettes to be tangled (and output compared) in R CMD check if they
    have a corresponding .Rout.save (thanks, @lentinj, #2018).

(mef)

doc: Updated devel/glib2 to 2.68.4nb1

(wiz)

glib2: getfsent() is not thread-safe.

Add patch from FreeBSD ports, via Sergio Lenzi in PR 56408.

Bump PKGREVISION.

(wiz)

doc: Updated math/R-tensorA to 0.36.2

(mef)

(math/R-tensorA) Uupdated 0.36.1 to 0.36.2, ChangeLog unknown

(mef)

doc: Updated math/R-stringdist to 0.9.8

(mef)

(math/R-stringdist) Updated 0.9.5.5 to 0.9.8

version 0.9.8
- Fixed some issues on C-level causing problems with the
  CLANG compiler. (Thanks to Brian Ripley for not only
  reporting this, but also sending updated code with
  fixes).

version 0.9.7
- Fixes in use of INTEGER() and VECTOR_ELT() after updates in R's C API.
  this affected 'afind' and 'max_length' (internally). (Thanks to Luke
  Tierny and Kurt Hornik for the notification).
- Fix in 'amatch' causing utf-8 characters to be ignored in some
  cases (thanks to Joan Mime for reporting #78).
- Fix: segfault when 'afind' was called with many search patterns or many
  texts to be searched.
- Fix: stringsimmatrix was not normalized correctly (Thanks to Tamas Ferenci
  for reporting GH).

version 0.9.6.3
- Resubmit. Fixed an URL redirect that was detected by CRAN.

version 0.9.6.2
- Resubmit. Fixed url issues detected by CRAN, added doi to description
  as per CRAN request.

version 0.9.6.1
- Bugfix: afind/grab/grabl returned wrong results on MacOS only.
  (thanks to Prof. Brian Ripley for the notification and for running tests
  on his personal machine and to Tomas Kalibera for making the
  ubuntu-rchk docker image available).

version 0.9.6
- New function 'afind': find approximate matches in text based on string distance.
- New functions 'grab', 'grabl': fuzzy matching equivalent to 'grep' and 'grepl'.
- New function 'extract': fuzzy matching equivalent of stringr::str_extract.
- New algorithm 'running_cosine': fast fuzzy text search using cosine distance.
- New function 'stringsimmatrix' (Thanks to Johannes Gruber).
- Number of threads used is now reported when loading 'stringdist'.
- Internal fixes (in some cases class() == 'class' was used).

(mef)

doc: Updated math/R-statmod to 1.4.36

(mef)

(math/R-statmod) Updated 1.4.32 to 1.4.36

25 Aug 2020: Statmod 1.4.35

- Fix Bug in tweedie(link.power=0) so that the resulting functions
  $linkinv() and $mu.eta() preserve the attributes of their
  arguments.

16 Feb 2020: statmod 1.4.34

- Improve the model description provided in the remlscoregamma() help
  page.

- tweedie() now checks whether `var.power` or `link.power` are
  character strings instead of numeric. If `var.power` is one of the
  standard family names ("gaussian", "poisson", "gamma" or
  "inverse.gaussian") or `link.power` is one of the standard link
  functions ("identity","log","inverse") then the argument is reset
  to the corresponding numerical value with a message, otherwise an
  informative error message is given.

- Cleaning up of internal code to avoid partial matching of function
  arguments, attributes or list component names. The automatic package
  tests are now run with the warnPartialMatchArgs,
  warnPartialMatchAttr and warnPartialMatchDollar options all set to
  TRUE.

4 Jan 2020: statmod 1.4.33

- The components returned by mixedModel2Fit() relating to fixed
  coefficients are now documented explicitly. The help page has been
  corrected to refer to the argument `only.varcomp` instead of
  `fixed.estimates`. The vector of `reml.residuals` is no longer
  part of the output.

- The test file has been slightly revised using zapsmall() so ensure
  that the test output file remains correct for R with ATLAS BLAS.

(mef)

doc: Updated math/R-splancs to 2.01.42

(mef)

(math/R-splancs) Updated 2.01.40 to 2.01.42, ChangeLog too old

(mef)

doc: Updated math/R-spacetime to 1.2.5

(mef)

(math/R-spacetime) Updated 1.2.4 to 1.2.5, ChangeLog too old

(mef)

(devel/py-ipykernel) fix build, +py-ipython_genutils for DEPENDS

(mef)

doc: Updated math/R-sandwich to 3.0.1

(mef)

(math/R-sandwich) Updated 2.5.1 to 3.0.1

# sandwich 3.0-1

* Extended the "Getting started" page with information on how to use _sandwich_ in
  combination with the _modelsummary_ package (Arel-Bundock) based on _broom_
  infrastructure (Robinson, Hayes, Couch). (Based on ideas from Grant McDermott.)
  <https://sandwich.R-Forge.R-project.org/articles/sandwich.html>

* Catch `NA` observations in `cluster` and/or `order.by` indexes for `vcovCL()`,
  `vcovBS()`, `vcovPL()`, and `vcovPC()`. Such missing observations cannot be
  handled in the covariance extractor functions but need to be addressed prior
  to fitting the model object, either by omitting these observations or by
  imputing the missing values. (Raised by Alexander Fischer on StackOverflow
  <https://stackoverflow.com/questions/64849935/clustered-standard-errors-and-missing-values>.)

* In `vcovHC()` if there are `estfun()` rows that are all zero and `type = "const"`,
  then the working residuals for `lm` and `glm` objects are obtained via
  `residuals()` rather than `estfun()`. (Prompted by an issue raised by
  Alex Torgovitsky.)

# sandwich 3.0-0

* Release of version 3.0-0 accompanying the publication of the paper
  "Various Versatile Variances: An Object-Oriented Implementation of
  Clustered Covariances in R." together with Susanne Koell and Nathaniel Graham
  in the _Journal of Statistcal Software_ at <https://doi.org/10.18637/jss.v095.i01>.
  The paper is also provided as a vignette in the package as
  `vignette("sandwich-CL", package = "sandwich")`.

* Improved or clarified notation in Equations 6, 9, 21, and 22 (based on
  feedback from Bettina Gruen).

* The documentation of the HC1 bias correction for clustered covariances
  in `vignette("sandwich-CL", package = "sandwich")` has been corrected (Equation 15).
  While both the code in `vcovCL()` and the corresponding documentation `?vcovCL`
  always correctly used (n-1)/(n-k), the vignette had incorrectly stated it as
  n/(n-k). (Reported by Yves Croissant.)

* The package is also accompanied by a `pkgdown` website on R-Forge now:
  <https://sandwich.R-Forge.R-project.org/>
  This essentially uses the previous content of the package (documentation,
  vignettes, NEWS) and just formatting was enhanced. But a few new features
  were also added:

  - A "Get started" vignette for the `pkgdown` page (but not shipped in the
    package) providing an introduction to the package and listing all
    variance-covariance functions provided with links to further details.
  - R/Markdown overview vignettes for the `pkgdown` page (but also not shipped
    in the package) linking the `Sweave`-based PDF vignettes so that they are
    easily accessible online.
  - A `README` with very brief overview for the `pkgdown` title page.
  - A nice logo, kindly provided by Reto Stauffer.

* All kernel weights functions in `kweights()` are made symmetric around zero now
  (suggested by Christoph Hanck). The quadratic spectral kernal is approximated
  by `exp(-c * x^2)` rather than `1` for very small `x`.

* In case the `Formula` namespace is loaded, warnings are suppressed now for
  processing formula specifications like `cluster = ~ id` in `expand.model.frame()`.
  Otherwise warnings may occur with the `|` separator in multi-part formulas with
  factors. (Reported by David Hugh-Jones.)

* The `bread()` method for `mlm` objects has been improved to also handle
  _weighted_ `mlm` objects. (Suggested by James Pustejovsky.)

(mef)

doc: Updated math/R-rstudioapi to 0.13

(mef)

(math/R-rstudioapi) Updated 0.11 to 0.13

# rstudioapi 0.13

* Fixed an issue where `rstudioapi::insertText()` would fail. (#208)

# rstudioapi 0.12

* Fixed an issue where remote `rstudioapi` calls would erroneously use
  a previous response in some cases.

* Allow `navigateToFile` to accept an empty file. This file will
  default to the file currently in view in the active column.

* Added `registerChunkExecCallback` and `unregisterChunkExecCallback`,
  used to execute a callback after a chunk is ran.

(mef)

Oops, array append syntax in Meson is +=. Should fix non-macOS builds.
Problem reported by wiz@.

(schmonz)

Actually make this MAKE_JOBS safe (hint from Joerg)

(martin)

doc: Updated games/flightgear to 2020.3.11

(nia)

doc: Updated games/simgear to 2020.3.11

(nia)

simgear, flightgear: 2020.3.11

Changes:
    Fix a crash introduced in 2020.3.10
    Improve error message when no aircraft search paths are available
    Fix a crash with misconfigured traffic taxi routes
    Update AI traffic schedules

(nia)

doc: Updated audio/pt2-clone to 1.34

(nia)

pt2-clone: add man page and desktop file, adapted from Debian

(nia)

pt2-clone: update to 1.34

v1.34 - 16.09.2021
- Bugfix: Pattern delay (EEx) did nothing on muted channels
- Bugfix: Less clicks/pops during E0x command (Set LED Filter)
- Added the missing Karplus-Strong command (E8x). Since this effect was
  sometimes replaced with visuals-syncing in demo music, it can be turned off by
  setting DISABE_E8X to TRUE in protracker.ini.
- Fix: Only do 2x oversampling if the audio output rate is below 96kHz. It
  would otherwise be a waste of CPU time (diminishing returns).
- Paula emulation is now slightly more accurate to a real Amiga (verified):
  1) Period/volume changes are now only updated when it's time to read
    a new sample point (period refetch). It was recently discovered that
    volume is also fetched (and cached) in the period refetch stage.
  2) During sample point fetching, the new sample point is now cached until it's
    time to read the next sample point. This can sometimes make a very small
    difference for sampledata-changing ProTracker effects (E8x / EFx).
- Removed BLEP'ing of volume changes, as it was no longer needed with the new
  way of doing volume changing, and it saves up some CPU time.

v1.33 - 04.09.2021
- The tracker now uses 2x oversampling and decimation for mixing and filtering,
  for (theoretically) less aliasing.
- Added Amiga 1200 low-pass filter (~34.4kHz, when not in A500 mode). This
  makes sense now, since we have twice the render/filter frequency than before.
- PAT2SMP now renders in higher quality thanks to the new high quality 2x
  decimator routine. Especially noticable in LO mode on high-frequency content.
- The "2x downsample" option while loading samples results in slightly less
  aliasing than before because of the new high quality 2x decimator.
- The MOD2WAV audio rate is now the same as the tracker's audio rate
- Stereo separation now works like it should. The volume is a bit louder than
  before if you increase the separation, but that's expected.
- Removed the hand-made cutoff tweak on the Amiga 500 low-pass filter. It now
  uses the correct nominal cutoff frequency. The RC filter routines have been
  replaced with new ones, and with these I didn't have to tweak the cutoff to
  make it sound similar to an A500 (probably related to filter prewarping?).

v1.32 - 12.08.2021
- Bugfix: If you had accidentally clicked somewhere on the sample data before
  using the "ramp volume" tool, it wouldn't work as expected.
- Bugfix: The BPM was a tiny fraction off because of an off-by-one error in
  the Amiga CIA period -> Hz calculation. The error was about ~0.02%, so it
  would only be measurable when comparing several minutes of recorded audio.

v1.31 - 19.06.2021
- The Disk Op. file sorting routine is now simplified and more ProTracker like
  (names are converted to uppercase before sorting). This affects the sorting.
- Bugfix: Don't allow mouse wheel scrolling while the volume/filter toolbox
  is open (sample editor).

v1.30 - 29.04.2021
- Bugfix: MOD2WAV would in some cases not render the correct amount of data
- Fixed an old PT bug where the pattern editor could sometimes show the next
  pattern's data one tick too early at the end of the pattern.
- Some small code refactoring

v1.29 - 14.03.2021
- Bugfix: Characters A..F were not accepted in the "Mix" text box in Edit Op. #3
- Screensaver is no longer disabled when the program is running

(nia)

textproc/ruby-nokogiri: fix dependency

Allow depending to misc/ruby-mini_portile2 2.7.0 and later.

* Override gemspec.
* There is dependency to mini_portile2 in ext/nokogiri/extconf.rb.  :(

No PKGREVISION bump since this change fix broken package.

(taca)

qjackctl: update DESCR

(nia)

doc: Updated audio/wmmixer to 1.9

(nia)

wmmixer: update to 1.9

Release 1.9  03 Mar 2021 Remove obsolete dynamic exception specification.

(nia)

Added devel/R-globals version 0.14.0

(mef)

(devel/R-globals) import R-globals-0.14.0

Identifies global ("unknown" or "free") objects in R expressions by
code inspection using various strategies (ordered, liberal, or
conservative).  The objective of this package is to make it as simple
as possible to identify global objects for the purpose of exporting
them in parallel, distributed compute environments.

(mef)

doc: Updated math/R-pixmap to 0.4.12

(mef)

(math/R-pixmap) Updated 0.4.11 to 0.4.12

Changes in pixmap version 0.4-12

  o Fixed some NAMESPACE problems

(mef)

doc: Updated math/nlopt to 2.6.2

(mef)

(math/nlopt)  Updated 2.4.2 to 2.6.2

# NLopt Release Notes

## NLopt 2.6.2

15 April 2020

* Fixed forced stop exception with dimension elimination ([#317])

* Fixed `get_initial_step` wrapping ([#319])

* Various build fixes ([#314], [#308], [#303], [#278])

## NLopt 2.6.1

13 April 2019

* Fix `nlopt_version` result for 2.6.x and update soname.

## NLopt 2.6

12 April 2019

* New `nlopt_set_upper_bound` and `nlopt_set_lower_bound` functions in the low-level C API to set one bound at a time ([#257]).

* There is no longer a separate `libnlopt_cxx` library: C++ algorithms (STOGO and AGS) are compiled and included by default ([#198]).

* Various build fixes ([#197], [#216], [#245], [#250], [#230], [#261], etc.), other fixes ([#242], [#258]).

## NLopt 2.5

26 July 2018

* New AGS global solver ([#194]), thanks to Vladislav Sovrasov.

* New `nlopt_get_numevals` function providing a built-in evaluation counter ([#160]).

* New `nlopt_get_errmsg` function for more descriptive error messages.

* Build system is converted to `cmake` ([#49]), thanks to Julien Schueller

* Plugins updated for recent Octave and Guile versions.

* Various other build fixes and minor bug fixes.

(mef)

doc: Updated math/R-nimble to 0.11.1

(mef)

(math/R-nimble) Updated 0.9.0 to 0.11.1

(from: inst/NEWS)
                            CHANGES IN VERSION 0.11.1 (May 2021)

USER LEVEL CHANGES

-- Add information about categorical sampler and univariate version of ESS
  sampler to `help(samplers)`.

BUG FIXES

-- Fix to the `posterior_predictive_branch` MCMC sampler, to update
  the log-probabilities of the sampled posterior predictive nodes (PR #1127).

                            CHANGES IN VERSION 0.11.0 (April 2021)

USER LEVEL CHANGES

-- Add new `posterior_predictive_branch` MCMC sampler, which is
  automatically assigned to trailing dependency node networks of entirely
  non-data nodes (jointly posterior predictive branches).  This sampler
  simulates jointly from the predictive distribtion of these posterior
  predictive node branches, and is designed to improve MCMC mixing of the
  branch, and consequently of the entire model (PR #1086).

-- Allow use of elliptical slice sampler for univariate nodes, which can be
  useful in multimodal problems (PR #1109).

-- Add a `getParents` method to the model API, allowing one to determine parent
  nodes, analogous to use of `getDependencies` to determine child nodes
  (PR #1094).

-- Add `getConditionallyIndependentSets` method (not yet documented) to the model
  API, allowing one to determine nodes that are conditionally independent of
  each other given parent nodes (PR #1094).

-- Improve efficiency of conjugate samplers by avoiding unneeded calculations
  when a conjugate relationship does not involve shifting or scaling (PR #1087).

-- Allow use of `nimNumeric`, `nimMatrix`, `nimArray` in model code (PR #1096).

-- Add progress bar to `getSamplesDPmeasure` (NCT issue 110).

-- Allow model definition using `if` without `else`, fixing a longstanding
  oversight (PR #1104).

-- Improve warning when multiple nodes provided to `getParam` (PR #1118).

-- Check during model building for unnamed elements of `data` and `inits`
  (PR #1117).

-- Remove error trapping to prevent use of variables in defining node names,
  such as `getDependencies('y[idx]')` as this is hard to check robustly and
  efficiently (PR #1122).

-- Improve error messages when reporting `getParam` cannot calculate a parameter
  when checking a model (PR #1112).

-- Error trap cases where model nodes are defined in two different declarations,
  adding check for overlapping multivariate nodes (PR #1110).

-- Improve error trapping of mis-formed stochastic declarations in models
  (PR #1106).

-- Increase maximum length of compiler output when using
  `compileNimble(..., showCompilerOutput = TRUE)` (NCT issue 205).

-- Point to parallelization example on r-nimble.org in relevant places of manual.

BUG FIXES

-- Fix a bug (issue #1091) causing incorrect node names when having more than
  100,000 elements in a vector node or in a dimension of a multi-dimensional
  node (PR #1092).

-- Fix `getNodeNames` to return no nodes when `latentOnly` is `TRUE` and model
  contains no latent nodes (PR #1116).

-- Fix checking for unknown nimbleFunction methods and improve related error
  trapping (PRs #1107, #1105).

DEVELOPER LEVEL CHANGES

-- Update our testing code/infrastructure to use latest testthat API (PR #1090).

-- Shift internal code to use `model$calculate(...)` style rather than
  `calculate(model, ...)` style for various node functions (PR #1114).

-- Clean up commented out code (PR #1098) and remove unused test files
  (PR #1097).

-- Update to a newer (but not latest) version of Eigen to suppress some compiler
  warnings (PR #1093).

                            CHANGES IN VERSION 0.10.1 (November 2020)

USER LEVEL CHANGES

-- Add `round` argument to `samplesSummary` (PR #1077).

-- `samplesSummary` function (and also `runMCMC(..., summary = TRUE)`) was made
  to be robust against non-valid values in posterior samples array (PR #1075).

BUG FIXES

-- Fix `makeParamInfo` when there is only one declID involved to address a bug
  affecting usage of `getParam`. This bug was introduced in version 0.10.0 when
  reducing memory use of `getParam` (PR #1016). This fixes incorrect behavior
  of conjugate samplers (because of incorrect inputs from `getParam`) under
  certain model structures, in particular state-space style models (PR #1080).

-- Prevent usage of marginal version of WAIC (i.e., when not monitoring all
  direct stochastic parents of data nodes); use of marginal version of WAIC in
  previous versions gave incorrect results (PR #1083).

DEVELOPER LEVEL CHANGES

-- Deprecate `samplerAssignmnentRules` system (PR #1078).

-- Deprecate `autoBlock` MCMC option (PR #1079).

                            CHANGES IN VERSION 0.10.0 (October 2020)

USER LEVEL CHANGES

-- Greatly extend BNP functionality with the CRP (Chinese restaurant process)
  distribution by allowing multiple observations to be grouped together (e.g.,
  for longitudinal or time series data) without requiring they be specified
  as a multivariate node (PR #1033).

-- Add a variety of conjugate cases to BNP conjugate samplers (PR #1033).

-- Greatly improve efficiency of model and MCMC building and configuration for
  BNP-based models with CRP components (PR #1033).

-- Move all sequential Monte Carlo (SMC; aka particle filtering) methods to
  new package `nimbleSMC`, including various particle-filter-based MCMC
  samplers.

-- Prevent use of variables in indexes of nodes, such as `y[idx]`, which was
  incorrectly being evaluated based on R scoping rules (PR #1064).

-- Allow use of `logdet` in model code (Issue #1018).

-- New `resetMV` argument available to `mcmc$run` method. In combination
  with `reset = FALSE`, specifying `resetMV = TRUE` will continue the current
  run of the MCMC, but discard any previously-collected samples
  (PR #1051; thanks to 'DJRP').

-- New methods `setMonitors` and `setMonitors2` added for MCMC configuration
  objects.  These methods replace the current set of monitors (or monitors2)
  with the specified variables (PR #1061).

-- Add `as.list` method for modelValues objects (PR #1060).

-- Update `getSamplesDPmeasure` function to improve efficiency and reduce
  output size; output is now a list of matrices (PR #1059).

-- Add `dimensions` argument to `nimbleMCMC` (PR #1058).

-- Add method `getWidthHistory` to slice sampler to retrieve sampling history
  information (PR #1057; thanks to 'rpatin').

-- Various improvements to the manual.

BUG FIXES

-- Fix a bug in k-fold cross-validation routine (`runCrossValidate`), where
  the merging of MCMC sampler configurations was done incorrectly and causing
  incorrect results (PR #1068).

-- Fix bug giving incorrect `dwish` density when using non-default S
  parameterization (PR #1017).

-- Fix incorrect NaN eigenvalues in singular normalized adjacency matrices
  under `dcar_normal` (PR #1019).

-- Update all MCMC sampler functions to use a new syntax for control list
  element extraction, which prevents a possible bug caused by R's partial
  matching of list names (PR #1065).

-- Define auto-generated simulation ('r') functions for user-defined
  distributions in the global environment to avoid scoping issues (PR #1063).

-- Update user-defined distribution processing so user-defined distributions
  can be defined inside functions (PR #1063).

-- Fix bug preventing use of `dirName` argument to `compileNimble` (PR #1062).

-- Fix a bug preventing model building when there are overly long names of
  model variables resulting from long deterministic expressions in model code
  (PR #1069).

-- Fix `buildMCEM` so it works with a compiled model as argument (PR #1028).

-- Fix `dmvt` so default unnamed parameters work (PR #1027).

-- Fix error in model building in corner case where
  `makeVertexNamesFromIndexArray2` made a simplifying assumption to conclude
  a block of nodes was contiguous (PR #1026).

-- Fix bug in `nimbleRcall` causing run-time warnings when `returnType` is void
  (PR #1013).

DEVELOPER LEVEL CHANGES

-- Improve efficiency of `getParam` implementation, which improves speed for
  MCMC compilation (PR #1016).

-- Improve MCMC sampling efficiency by not copying data nodes, only data node
  logProbs, during sampler execution for various samplers (PR #1040).

-- Update Travis testing to use R 4.0.

-- Remove `GID_map` internal to modelValues (PR #1032).

-- Remove deprecated function `getLoadingNamespace` and (deprecated) use of
  `where=getLoadingNamespace`. Also improve handling of environments set up
  by `nimbleFunction` to make it easier to write packages depending on NIMBLE
  (PRs #1029, 1011).

-- Force intermediates of index range expressions to be of type 'int' for use
  in AD (PR #1024).

                            CHANGES IN VERSION 0.9.1 (May 2020)

USER LEVEL CHANGES

-- Switched from use of `system` to `system2` to avoid problems with installation
  under R 4.0 on Windows (PR #1003).

-- Modify various adaptive MCMC samplers so the exponent controlling the scale
  decay of the adaptation is adjustable by user (rather than hard-coded at 0.8
  (PR #981).

-- Allow `pmin` and `pmax` to be used in models (PR #982).

-- Add documentation for `is.na`,`is.nan`,`any`,`all` (PR #988)

-- Add system option `MCMCuseConugacy` to control whether conjugate samplers are
  used (PR #998).

-- Adds checks for `niter`, `nburnin` in the `mcmc$run` method (PR #980).

-- Modify print handling in `addSampler` and `configureMCMC` (PRs #986, 989).

-- Improve handling of NA values in `dCRP` to avoid error messages when building
  models (PR #994).

-- Avoid monitoring top-level data nodes in models (PR #1006).

BUG FIXES

-- Modify MCMC `autoBlock` routine to only group Wishart, Inverse-Wishart, and
  Dirichlet nodes with themselves, to avoid violating the constraints of those
  nodes (PR #999).

-- Fix incorrect error message from `warnRHSonlyDynIdx` when variable appears
  multiple times on right-hand side of a model expression (PR #997).

-- Fix `checkDistributionFunctions` to respect default `nDim=0` when extracting
  first argument, to avoid error when dimension not specified in user-defined
  distributions (PR #992).

-- Fix `print` option of `addSampler` (PR #986).

-- Improve handling of cases where indexing goes beyond extent of variable in
  `expandNodeNames` and related queries of model structure (PR #977).

DEVELOPER LEVEL CHANGES

-- Use `inherits` rather than testing for equality of `class(object)` (PR #988).

(mef)