• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1[/
2            Copyright Hans Dembinski 2018 - 2019.
3   Distributed under the Boost Software License, Version 1.0.
4      (See accompanying file LICENSE_1_0.txt or copy at
5            https://www.boost.org/LICENSE_1_0.txt)
6]
7
8[section:history Revision history]
9
10[heading Boost 1.72]
11
12* New features
13  * Better deduction guides for axis types and histogram type on C++17 compilers
14  * performance improvements in the indexing code
15  * new histogram::fill method for accelerated filling from chunks of values
16  * ASCII bar charts for 1D histograms when boost/histogram/ostream.hpp is included (contributed by Przemyslaw Bartosik)
17  * Passing invalid axis options causes user-friendly compile-time errors
18  * Wrong usage of weight() and sample() causes user-friendly compile-time errors
19  * algorithm::empty returns true if all histogram values are equal to the default value (contributed by Henry Schreiner)
20  * algorithm::reduce with shrink is now well defined when values are not bin edges
21  * axis::traits::rank returns the number of values that this axis accepts
22  * axis::traits::is_continuous returns whether an axis is continuous or discrete
23  * axis::traits::is_inclusive returns whether an axis has a bin for any possible input
24  * limited weight support for mean accumulator
25  * accumulators::weighted_mean::sum_of_weights_squared method added
26
27* Fixes
28  * Support of -std=c++2a and -stdlib=libc++ on clang, compatibility with gcc-9
29  * Fixed: weight could be passed instead of a sample during filling without triggering an error
30  * Fixed: segfault when too many indices were passed to algorithm::project
31  * Fixed: indexed range generator did not work with storage based on std::array
32  * Fixed: weighted_mean() + weighted_mean() != weighted_mean() (discovered and reported by Henry Schreiner)
33  * Fixed: axis::option::test(...) returned true if any bits in the test mask were set (now returns true if all bits in the test mask are set)
34  * Protecting calls to min, max with against macro expansion
35  * Replaced all crlf with lf in concepts.qbk
36  * Missing throw of std::invalid_argument when user passes too many axes to histogram
37  * Corrected documentation of serialization support in concepts section
38  * Resolved warnings for various compiler versions
39
40* Other
41  * Added Boost.Histogram logo
42  * Added missing copyright notices
43  * axis::category::value returns copy for scalar types and const reference otherwise
44  * std::ostringstream not used anymore to generate exception messages to reduces code bloat
45  * Documentation improvements
46    * Restructured overview section in documentation
47    * Updated user guide to demonstrate new features
48    * Updated accumulator examples
49    * Concepts explain how accumulators can optionally accept weights
50  * Updated benchmark code
51  * New test checks consistency of b2 and cmake build systems
52  * New test checks One-Definition-Rule (fails if non-templated function is not declared inline)
53  * Serialization code decoupled from Boost.Serialization
54  * Removed dependency on Boost.CallableTraits
55
56[heading Boost 1.71]
57
58* New features
59  * Thread-safe accumulators boost::histogram::accumulators::thread_safe based on std::atomics
60  * Support for thread-safe storages
61  * Support for compiling without exceptions/RTTI (increases performance by 10-20 %) [with contributions from Glen Fernandez]
62  * Performance improvements for 1D and 2D histograms
63  * boost::histogram::indexed now returns forward iterator range instead of input iterator range
64  * boost::histogram::indexed_range::accessor is now non-copyable and acts like reference to cell value, making most algorithms from the stdlib work
65  * boost::histogram::algorithm::reduce
66    * New slice option
67    * Fuse shrink, slice, and rebin options passed for the same axis
68    * Support histograms with some axis types with reduction support
69  * boost::histogram::algorithm::project accepts runtime indices for static histograms
70
71* Fixes
72  * boost::histogram::algorithm::reduce also works on histograms that have some axis types without reduction support
73  * boost::histogram::axis::traits::update now works correctly for boost::histogram::axis::variant
74
75* Other
76  * 100 % test coverage
77  * Drastically reduced internal Boost dependencies
78  * Improved documentation and examples
79  * Internally replaced boost::variant with boost::variant2
80  * boost::histogram::axis::traits::is_reducible detects reducible axis types
81  * Cleanup and refactoring of internals
82  * Guarantee no-throw moves for all builtin axis types
83  * Improved internal benchmarks
84  * Compile cleanly at higher warning levels
85
86[heading Boost 1.70]
87
88First Boost release, version 4.0 in former internal counting.
89
90* Removed Python bindings, will be developed in separate repository
91* All axes can be made optionally circular, except category axis
92* All axes now support generic attached metadata
93* All axes have now independently configurable underflow/overflow extra bins
94* Removed circular axis (which is just a circular regular axis)
95* Added indexed adaptor generator for convenient and fast iteration over histograms
96* Support for axes that can grow in range
97* Support for axes which accept multiple values (example: hexagonal binning)
98* Support for profiles and more generally, arbitrary accumulators in each cell
99* Support for serializing to xml archives
100* Added compatibility with Boost.Range, Boost.Units, and Boost.Accumulators
101* Added deduction guides for axis types and histogram
102* Performance improvements
103* Renamed `adaptive_storage` to `unlimited_storage`
104* Replaced `boost::multiprecision::cpp_int` with custom type to decouple libraries
105* Internal simplification of `axis::variant`
106* Make all storages behave more like containers, simplifying the design
107* Histograms supports add, sub, mul, div
108* replaced `operator[]` in axis with explicit method `bin`
109* replaced `operator()` in axis with explicit method `index`
110* replaced internal use of `boost::containers` with stdlib containers
111* Much improved docs, reference documentation, user guide, more examples
112
113[heading 3.2 (not in Boost)]
114
115* Allocator support everywhere
116* Internal refactoring
117
118[heading 3.1 (not in Boost)]
119
120* Renamed `bincount` method to `size`
121* Support for axes with only overflow and no underflow bin
122* category axis now by default has bin for "other" input that does not fall
123  into the predefined categories, making it consistent with other axes
124* NaN is now consistently put into overflow bin for all axes
125* Eliminated warnings about safe internal conversions on MSVC
126* Established a cpp house style with corresponding .clang-format file
127* Better detection of Python library on all systems
128* Improved code coverage by testing more input errors
129* Raise ValueError instead of generic RuntimeError in Python on input errors
130
131[heading 3.0 (not in Boost)]
132
133* Support for efficient adding of multiple histograms and scaling
134* Re-design of category axis as a general mapping between unique values and bins
135* Re-design of the bin description an axis returns upon element access
136* Interface cleanup and simplification, more consistency between Python and C++ interface
137* Iterators for histogram and axes for STL compatibility
138* Regular axis in C++: Allowing transforms with state
139* Regular axis in Python: Support for all C++ transforms
140* Polymorphic axis::any type
141* Internal refactoring from boost::mpl and boost::fusion to std::tuple and boost::mp11
142* Windows support
143
144[heading 2.0 (not in Boost)]
145
146* Added static_histogram (v1.0 only had dynamic_histogram).
147* Merged wfill(...) and fill(...) interface.
148* Support custom allocators in storage classes.
149* Replaced static_storage with array_storage.
150* Replaced dynamic_storage with unlimited_storage, which adds the capability to grow the bin counter into a cpp_int, thus avoiding integer overflow completely.
151* Serialization uses binary_archive instead of text_archive. The latter is portable, but the performance is terrible.
152* Python interface changed: histograms are now iterable, returning axis classes
153* Support reduction (removing a subset of axes and returning the equivalent histogram; in other words, those axes are integrated over)
154
155[heading 1.0 (not in Boost)]
156
157* First stable version.
158
159[endsect]
160