v0.3.0

New features:

• Nuisance parameters for ratio-based methods! This required a major refactoring of madminer.sampling.
• New madminer.limits submodule that calculates expected and observed p-values.
• SALLY can profile the estimated score over nuisance parameters (then its output is n-dimensional rather than (n+k)-dimensional, where n is the number of parameters of interest and k is the number of nuisance parameters).
• New madminer.analysis submodule with a generic DataAnalyzer class that can be used to access the general setup, weighted events, and cross sections. The classes SampleAugmenter, FisherInformation and AsymptoticLimits subclass it, leading to a more unified interface.
• Sampling speed-up with parallelization when using n_processes>1 for any sampling call.

Breaking / API changes:

• New file format for trained neural networks (only the *.json files are different) -- old models won't load
• The one-size-fits-all class MLForge is replaced by four different classes ParameterizedRatioEstimator, DoubleParameterizedRatioEstimator, LikelihoodEstimator, and ScoreEstimator.
• EnsembleForge is now Ensemble, with renamed functions and less clutter (at the cost of some unimportant functionality)
• Renaming: DelphesProcessor -> DelphesReader, LHEProcessor -> LHEReader
• madminer.morphing now lives in madminer.utils.morphing, and the Morpher class was renamed PhysicsMorpher (since we also have a NuisanceMorpher class)
• The madminer.sampling has changed in a number of ways. The high-level functions have been renamed: extract_samples_train_plain() -> sample_train_plain(), extract_samples_train_local() -> sample_train_local(), extract_samples_train_global() -> train_samples_density(), extract_samples_train_ratio() -> sample_train_ratio(), extract_samples_train_more_ratios() -> sample_train_more_ratios(), extract_samples_test() -> sample_test(), extract_cross_sections() -> cross_sections(). In addition to the physical parameters theta, they all now take descriptions of the nuisance parameters nu as input argument, for which new helper functions exist.
• The helper functions in madminer.sampling are also different: constant_benchmark_theta() -> benchmark(), multiple_benchmark_thetas() -> benchmarks(), constant_morphing_theta() -> morphing_point(), multiple_morphing_thetas() -> morphing_points, random_morphing_thetas() -> random_morphing_points()
• New default evaluation split of 0.2 (so with the defaults, there will be 60% train, 20% validation, 20% test events)

Internal changes:

• madminer.ml largely rewritten
• madminer.sampling largely rewritten
• madminere.utils.analysis got removed, the old functions are now split between madminer.analysis and madminer.sampling