gpt/NEWS at master · danheck/gpt · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
gpt 0.6.1 (November 2023)
===========

* fix test_fit() issue due to incompatible S3 class cdf("gpt")
* updated citation("gpt") to pubished Psychometrika paper (Heck et al., 2018)


gpt 0.6.0 (January 2018)
===========

* Option to use different continuous distribution across component distributions via fit.gpt(..., latent="custom")
* Renamed functions: gpt_fit, gpt_gen, gpt_gen_sample, plot.gpt_fit, hist.gpt_fit, select_gpt, test_nested, test_fit
* Renamed S3 class to gpt_fit (instead of gpt.fit)
* New function update.gpt_fit() to fit nested models


gpt 0.5.0 (December 2017)
===========

* Goodness-of-fit statistics
* Improved stability and bug fixes


gpt 0.3.0 (August 2016)
===========

* testthat routines for unit checks
* Predict MPT probability/mean/quantiles/density: predict.gpt.fit (S4 method)
* Plot model fit: cdf, qplot
* Subsetting to plot model fit separately per group


gpt 0.2.0 (June 2016)
===========

* S4 classes for mpt, gpt, univariate distributions
* S4 methods for densities and data generation (mpt, gpt, univariate distributions)
* better starting values, separately for each continuous distribution
* lower and upper bounds stored internally and adjusted for each eta parameter


gpt 0.1.0 (June 2015)
===========

* first version of gpt
* data generation: gen.gpt, gen.gpt.sample (allowing for heterogeneity)
* model fitting: fit.gpt (single data sets + whole sample using parallel computing)
* plotting model fit: plot.gpt (histgrams separately for response categories); qplot.gpt (x-axis: response probabilities per tree, y-axis: continuous quantiles)
* model comparisons by AIC, BIC: gpt.select
* latent continuous distributions: normal, lognormal, wald, gamma, exGauss, weibull