v0.4.5

update version

v0.4.4

Commits

• [17c691d]: try rescale always (giadasp)

v0.3.0

Updates

• Evaluate abilities only for examinees which have latent.assessed == false
• Estimate parameters only for items which have parameters.calibrated == false
• Fix bugs in information functions
• Add bootstrap for MMLE estimation

v0.2.0

Updates

• Add references for the PolyaGamma sampler
• Refactor estimation methods
• Add Maximum Marginal Likelihood Estimation (MMLE) method

v0.1.5

Updates

• Major code refactoring

v0.1.4

Updates

• Methods for AbstractParameters and AbstractLatents, such as probability and expected_information_item, are now internals, i.e. they are not exported by the module and their names are preceded by _.

v0.1.3

Updates

• Items and examinees types have been refactored. Item1PL, Item2PL, Item3PL, Examinee1D types have been substituted by Item and Examinee, respectively. The type of the item now depends only on its item parameters (Parameters1PL, Parameters2PL or Parameters3PL).
• Docs have been revised and pages "Bayesian IRT" and "Distributions" have been added. Docs are still a work in progress.

v0.1.2

Updates

• Add abstract subtypes for dichotomous (binary) responses and item parameters.

v0.1.1

new version

v0.1.0

fix bug