typo: node -> infoset in _resolve_infoset#535
Merged
rahulsavani merged 114 commits intomaint16_2from Jul 17, 2025
Merged
Conversation
This refactors the build of the pygambit extension to create more granular libraries of the C++ code, with the aim of reducing unnecessary re-builds when changes are made.
This adds maximum-likelihood estimation of agent logit QRE, in parallel to the methods for strategic logit QRE. Function names for estimation for the strategic version have been modified to give a parallel naming scheme for both.
This is a substantial refactoring and cleanup of path-following and the interface to QRE tracing. The only new functionality is to implement the ability to stop at an interior local maximiser when fitting QRE. Otherwise, other changes are internal to simplify calling the various QRE-related routines and separate concerns (e.g. formatting/printing is no longer part of the QRE tracer itself but properly delegated to a provided observer function if that is desired).
Bumps [jidicula/clang-format-action](https://github.com/jidicula/clang-format-action) from 4.11.0 to 4.12.0. - [Release notes](https://github.com/jidicula/clang-format-action/releases) - [Commits](jidicula/clang-format-action@v4.11.0...v4.12.0) --- updated-dependencies: - dependency-name: jidicula/clang-format-action dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com>
# Conflicts: # ChangeLog
Bumps [jidicula/clang-format-action](https://github.com/jidicula/clang-format-action) from 4.12.0 to 4.13.0. - [Release notes](https://github.com/jidicula/clang-format-action/releases) - [Commits](jidicula/clang-format-action@v4.12.0...v4.13.0) --- updated-dependencies: - dependency-name: jidicula/clang-format-action dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com>
* Bump actions/upload-artifact from 3 to 4 Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 3 to 4. - [Release notes](https://github.com/actions/upload-artifact/releases) - [Commits](actions/upload-artifact@v3...v4) --- updated-dependencies: - dependency-name: actions/upload-artifact dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com> * Update action for building executables to upload-artifact@v4 --------- Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Ted Turocy <ted.turocy@gmail.com>
* In estimating using the empirical payoff approach: * Return the estimated strategy profiles instead of a list of probabilities. * Improve the calculation of log-probabilities internally for better numerical performance, especially with small probabilities. * In estimating using the equilibrium correspondence, expose stepsize control parameters in parallel to other correspondence-based methods.
…ltiple lambda values in the same invocation. Closes #332.
* Simplify and rationalize naming of various functions to work with QRE correspondence * Move functions other than logit_solve into the qre module. * Added an appropriate section to the user guide summarizing QRE facilities.
This cleans up the implementations of the enumeration of possible Nash subsupports: * The principal enumeration methods are extracted from enumpoly into their own directory. * Unused or obsolete code has been removed, and some aspects of the implementation have been refactored for simplification. * Enumeration for the strategic game has been experimentally exposed in pygambit.
* Refactors and modernises the internal implementaton of enumpoly for strategic games * The heuristic-search version of enumpoly now uses the same per-support solver as the full enumeration
This drops support for Python 3.8. In addition, * Updates pre-commit actions * Updates Python style for 3.9/3.10 now that 3.8 is not supported.
# Conflicts: # .pre-commit-config.yaml
Exports the reduced strategic representation of a game as a collection of numpy arrays. Closes #461.
This is a substantial reworking of the implementation of the sequence form concept currently inside enumpoly. This significantly simplifies the implementation through a combination of better organisation and the use of modern C++ features which did not exist when this implementation was first done. In due course a more polished version of this should become part of the game-representation library. This is progress towards that, but is not yet fully stabilised.
This builds on previous work to bring the implementation of enumpoly on extensive forms (via the sequence form) up-to-date. This is a substantial refactoring that should improve the readability and maintainability of these routines. Further, the `GameSequenceForm` class has now moved towards being ready to move to be part of the game representation library, which the goal is to do in due course.
This does some cleanup and updating of behavextend. This involves no functionality changes, but removes unneeded/duplicate code, and uses modern features like STL containers and range-based for-loops where it is straightforward to do so. This is not a complete modernisation as there are a few places where modernisation is not just a trivial reworking. In addition, it is not at all clear that these functions are necessarily fully correct, or that the task they are solving could not be handled more easily!
This reimplements our support enumeration algorithm for strategic games, using the approach proposed by Porter, Nudelman and Shoham (2004). An experimental version of this had been previously included, written by Litao Wei. This version is re-built from the ground up as a replacement. Further, we now use this approach as the approach for generating supports, and have removed the old method entirely, as there is no real benefit to the way we used to do it over what PNS propose.
This implements the stopAfter argument to enumpoly for both strategic and extensive games, and makes it available in the command-line tool and in pygambit.
Parsing of .nfg files has heretofore been somewhat tolerant. If too many payoffs or outcomes were given, extra ones were ignored silently; if not enough were given, the missing ones were set to zero. With this, checks are now done to ensure the correct number of outcomes or payoffs (as appropriate) are provided. Closes #119.
…ve` and related functions Closes #459.
This does some general cleanups of routines to read and write savefiles: * Creates a standardised ReadXXXFile/WriteXXXFile set of functions. * Uses iterators to make the code, specifically the writing code, more generic. * Removes uses of Gambit containers (and manually-managed linked lists!) in favour of STL equivalents.
This extends `enumpoly_solve` with the option to use `PHCpack` [1]_ to solve the systems of polynomial equations. Closes #165. .. [1] https://homepages.math.uic.edu/~jan/PHCpack/phcpack.html
Improve organisation of code in core/: * Remove header dependency on game classes * Update file paths in copyright notice * Update #include guards to match file paths
Adds an internal count of the number of nodes in a game tree, rather than counting nodes explicitly when asked. Correspondingly, in Python, `nodes` is now a member (returning a shim object) rather than a function.
* Eliminates derived `GameTreeInfosetRep` and `GameTreeNodeRep` classes. * Moves operations which make strucutral changes to game trees to `GameTreeRep` rather than individual object classes.
This implements a new template class, ElementCollection, which represents a collection of objects in a game. This handles efficient access and iteration over game objects without the need to allocate a new container. Closes #516
This provides computation of the set of plays (terminal nodes) consistent with passing through a specified node, information set, or action: * `GameRep.GetPlays()` in C++ * Properties `Node.plays`, `Infoset.plays`, `Action,plays` in Python. Closes #517.
This provides an interface to the mapping from strategy to actions in an extensive game: * `GameStrategyRep::GetAction(GameInfoset)` in C++ * `Strategy.action(Infoset)` in Python`
This implements a Nodes collection class in a game, representing the set of all nodes in the game tree (if defined). The default iteration of the set of nodes is defined as depth-first traversal. Closes #530.
Member
|
This has been a bug since 16.2, so we should do the PR against maint16_2. |
tturocy
requested changes
Jul 17, 2025
Member
tturocy
left a comment
tturocy
left a comment
There was a problem hiding this comment.
This is fine, but we need to have this against the branch maint16_2 instead of master.
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
5 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.
No description provided.