Poisson geometric#1
Closed
JohanMabille wants to merge 25 commits into
Closed
Conversation
…ntainer (xtensor-stack#2880) # Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description The `index_mapper` class provides functionality to convert indices from a view's coordinate system to the corresponding indices in the underlying container. This is particularly useful for views that contain integral slices (fixed indices), as these slices reduce the dimensionality of the view. Example: ```cpp xt::xarray<double> a = xt::arange(24).reshape({2, 3, 4}); auto view1 = xt::view(a, 1, xt::all(), xt::all()); // Fixed first dimension index_mapper<decltype(view1)> mapper; // Map view indices (i,j) to container indices (1,i,j) double val = mapper.map(a, view1, 0, 0); // Returns a(1, 0, 0) double val2 = mapper.map(a, view1, 1, 2); // Returns a(1, 1, 2) ``` --------- Co-authored-by: Alexandre Hoffmann <alexandre.hoffmann@polytechnique.edu>
# Checklist - [ ] The title and commit message(s) are descriptive. - [ ] Small commits made to fix your PR have been squashed to avoid history pollution. - [ ] Tests have been added for new features or bug fixes. - [ ] API of new functions and classes are documented. # Description <!--- Give any relevant description here. If your PR fixes an issue, please include "Fixes #ISSUE" (substituting the relevant issue ID). -->
…or-stack#2889) # Checklist - [x] The title and commit message(s) are descriptive. - [ ] Small commits made to fix your PR have been squashed to avoid history pollution. - [ ] Tests have been added for new features or bug fixes. - [ ] ~~API of new functions and classes are documented.~~ # Description This PR fixes an error pertaining to reading the length of the header of an npy file. The header is read as two `char`s and then combined into a `uint16_t`. Previously due to `char` being a signed value, any value with the MSB set would get converted as if it were a negative number and thus misinterpreted. For example, the bytes `F6 00` which means a header of 246 bytes would be interpreted as a header of 65526 bytes. I think the code could be shortened a little, but I think when messing with bits like this it's best to be explicit. Also, the compiler will likely optimize it anyway.
…k#2891) # Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description The goal of this pullrequest is to fix thee error and warning during the documentation generation. --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com>
…tensor-stack#2871) (xtensor-stack#2882) This PR addresses GitHub issue xtensor-stack#2871 where using auto with xtensor reducer functions (like amax, sum with keep_dims) causes crashes or incorrect results in optimized builds, particularly on Android ARM64. **Root Cause:** Lazy expressions hold references to local variables. When intermediate results are stored with auto, these references become dangling when the function returns, leading to undefined behavior. **Changes:** - docs/source/pitfall.rst: Added detailed warning section explaining the issue and providing three solutions: 1. Use explicit container types (xt::xtensor<T, N>) 2. Use xt::eval() to force immediate evaluation 3. Use evaluation_strategy::immediate with reducers - test/test_xmath.cpp: Added test demonstrating correct patterns **Testing:** - All 83 xmath tests pass - All 64 xreducer tests pass Fixes xtensor-stack#2871
Commit generated through diskarzhan. With ci added to prevent regression.
This workaround is for old compilers not supported anymore. Edit: `XTENSOR_CONSTEXPR_ENHANCED_STATIC` is still required on Windows.
# Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description This PR fixes incorrect index mapping in xview assignments when a view starts with one or more newaxis() slices. The change updates the index computation to use a newaxis-aware slice index before applying integral-slice adjustments. That keeps writes through views with multiple leading newaxis() entries aligned with the correct element in the underlying tensor. Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com>
# Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description Add clang 22 and gcc 15 in the CI --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com> Co-authored-by: Copilot <copilot@github.com>
…stack#2909) # Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description Address xtensor-stack#2860 Now `clips` follow the rule: When a_min is greater than a_max, clip returns an array in which all values are equal to a_max --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com> Co-authored-by: Copilot <copilot@github.com>
# Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description Add Linux ARM build on the CI --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com>
# Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description xcsv_config is now used to configure dump csv function --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com> Co-authored-by: Copilot <copilot@github.com>
…h mismatched rank (xtensor-stack#2907) # Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description Addressing xtensor-stack#2792 Assigning a higher-dimensional expression to a fixed-rank [xtensor<T, N>] caused a buffer overflow (stack smashing) in release builds: ```c++ auto a = xt::xtensor<float, 1>::from_shape({2}); a = xt::expand_dims(a, 0); // expand_dims returns 2D view, shape {1, 2} // buffer overflow: resize truncates shape {1,2} -> {1}, then copies 2 elements into 1 element buffer ``` --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com>
# Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description Fix `xt::masked_view` pretty-printing when the underlying data is non-optional. Related to: xtensor-stack#2495 The generic non-fundamental printer was streaming masked proxy values directly, which could lose the correct masked/value state during formatting. This change keeps the existing generic printer path but handles `xtl::xmasked_value` with a compile-time branch and streams a stabilized value via unary `+`. ## Changes - include `xtl/xmasked_value_meta.hpp` in `xio.hpp` - update the generic non-fundamental printer to use `if constexpr` - stream masked values with `buf << +val` - add a regression test for `masked_view` stream output on plain `xarray` data --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com>
# Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description Address xtensor-stack#2867 Changed `xtl::is_integral<T>::value` to `xtl::is_integral<std::decay_t<T>>::value` in the `drop()` function template. When passing a variable like `size_t index = 1`, `T` deduces to `unsigned long&`, a reference type, which `is_integral` rejects. Decaying strips the reference so the integral branch is correctly taken. --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com> Co-authored-by: Johan Mabille <johan.mabille@gmail.com>
# Checklist - [ ] The title and commit message(s) are descriptive. - [ ] Small commits made to fix your PR have been squashed to avoid history pollution. - [ ] Tests have been added for new features or bug fixes. - [ ] API of new functions and classes are documented. # Description <!--- Give any relevant description here. If your PR fixes an issue, please include "Fixes #ISSUE" (substituting the relevant issue ID). --> --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com>
# Checklist - [x] The title and commit message(s) are descriptive. - [x] Small commits made to fix your PR have been squashed to avoid history pollution. - [x] Tests have been added for new features or bug fixes. - [x] API of new functions and classes are documented. # Description Fix xt::vstack so it works with xtensor_fixed inputs when stacking a mix of 1-D and 2-D fixed-shape expressions. Before this change, vstack could try to build a runtime shape for fixed-shape inputs, which fails for cases like stacking xshape<1, 2> with xshape<2, 2>. The fixed-shape path now computes the stacked result shape at compile time instead. --------- Co-authored-by: Alexis Placet <2400067+Alex-PLACET@users.noreply.github.com> Co-authored-by: Copilot <copilot@github.com>
Author
|
Wrong target, sorry for the noise. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Cherry-pick of last commit from xtensor-stack#2893, easier to do so than rebasing.