v0.1.5

Added

• remove methods to open data into bins or filtered and replace with method on Dataset
• wrap Events inside Datasets in Arc to reduce bootstrap copying
• add benchmark for opening datasets
• add method to resample datasets (bootstrapping)

Other

• switch to Codspeed for benchmarking
• update plot and add output txt file for example_1 and reorganize directory structure
• refactor data loading code into a shared function

v0.1.4

Added

• add gen_amp config file for Python example_1
• add python example
• add Debug derive for Parameters
• add method to input beam polarization info and assume unity weights if none are provided
• adds a LikelihoodScalar term that can be used to scale LikelihoodTerms by a scalar-valued parameter
• expose the underlying dataset and Monte-Carlo dataset in the Python API for NLL and add method to turn an NLL into a LikelihoodTerm
• some edits to convert module and exposure of the convert_from_amptools method in the main python package
• add gradient calculations at Amplitude level
• add amptools-to-laddu conversion script to python package
• add python API for likelihood terms and document Rust API
• proof-of-concept for Likelihood terms
• put Resources in Evaluator behind an Arc<RwLock<T>>
• Add LikelihoodTerm trait and implement it for NLL

Fixed

• update example_1.py to allow running from any directory
• change NLL implementation to properly weight the contribution from MC
• properly handle summations in NLL
• correct type hints
• ensure extension-module is used with the python feature
• make sure rayon-free build works
• these indices were backwards
• this should correctly reorganize the gradient vectors to all have the same length
• correct some signatures and fix PyObserver implementation

Other

• some stylistic changes to the README
• update README.md to include the first python example
• remove lints
• move kwarg extractor to be near parser
• update ganesh to latest version (better default epsilons)
• move parsing of minimizer options to a dedicated function to reduce code duplication
• add sample size specification
• move Likelihood-related code to new likelihoods module
• change benchmark config
• store Expressions inside Evaluators to simplify call signatures

v0.1.3

Added

• add options to the minimization callables and add binned Dataset loading to Python API
• add filtered and binned loading for Datasets
• export Status and Bound structs from ganesh as PyO3 objects and update minimize method accordingly
• add Debug derive for ParameterID
• add LadduError struct and work in proper error forwarding for reading data and registering Amplitudes
• use AsRef generics to allow more versatile Variable construction
• add ganesh integration via L-BFGS-B algorithm
• update to latest PyO3 version

Fixed

• missed one fully qualified path
• correct some namespace paths
• add Dataset and Event to variables
• add scalar-like Amplitudes to python namespace
• reorder expression and parameters
• remove main.rs from tracking

Other

• update minimization example in README.md
• fix doctest
• update ganesh version
• switch order of expression and parameters in evaluate and project methods

v0.1.2

Other

• remove tag check

v0.1.0

chore: Release package laddu version 0.1.0