-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathNEWS
96 lines (65 loc) · 3.02 KB
/
NEWS
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
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
metaBMA 0.6.9
===========
* Updated syntax for Stan arrays (thanks to @andrjohns)
* Dependency on newest version of rstan (>= 2.26.0)
* Updated references for tutorials on meta-analysis with model averaging
* Bug fixes for passing of arguments in plot_forest() (#23 and #24)
metaBMA 0.6.7
===========
* Transformation of effect sizes via transform_es()
* Sensitivity analysis via meta_sensitivity()
* New prior family: gamma distribution
* Scale of default prior for log odds ratio in meta_default() changed using the
scaling factor pi/sqrt(3) = 1.81 instead of 2.00 (thanks to Frantisek Bartos)
metaBMA 0.6.6
===========
* Upgrade to new rstantools folder structure
* New default prior for effect size: Cauchy distribution with scale=0.707 instead of normal distribution with scale=0.3
* New defaults in meta_default() wrapper function
* Faster CRAN checks and tests
metaBMA 0.6.4
===========
* bugfix: argument "ci" for credibility intervals not working
* reference added: Gronau et al. (2020)
metaBMA 0.6.3
===========
* increased stability and precision of model-averaged posterior distribution and estimates (based on density approximations)
metaBMA 0.6.2
===========
* Moderator analysis: rename slope parameters "alpha" to "beta"
* Bugfix for meta_bma(): Only use H0 models for averaging of "d" parameter
* New tests: Scheibehenne (2017)
metaBMA 0.6.0
===========
* new function meta_ordered() for order-constrained study-effects in random-effects meta-analysis
* table with estimates shows convergence statistics (Rhat, n_eff)
* meta_default(): new labels for effect = "d", "r", "z", "logOR"
* minor bugfixes and improvements
metaBMA 0.5.0
===========
* Major refactoring (breaks compatibiltiy with previous versions)
* Possible to provide data frame via argument 'data'
* Removed arguments "d.par" and "tau.par": priors are now defined via d=prior(...), tau=prior(...)
* Possibility to fit random and fixed effects meta-analysis with moderators in stan (with JZS priors)
* Computation of log marginal likelihood with Stan samples and bridge sampling (via logml="stan")
* Improved numerical integration via integrate() [posterior distribution shifted to zero]
metaBMA 0.3.9
===========
* Updated citation for CRAN
* Added examples for meta_bma() and meta_random()
* Minor bug fixes
metaBMA 0.3.8
===========
* Data sets 'power_pose' and 'power_pose_unfamiliar' added
* Data set 'facial_feedback' added
* More informative description file
* Requirements for CRAN
metaBMA 0.3.0
===========
* First stable version
* High-level functions meta_bma() and meta_default() perform model averaging for standard models (fixed, random + H0, H1)
* Plotting functions for averaged/random-effects/fixed-effects meta-analysis via plot_forest() and plot_posterior()
* Meta-analysis models are fitted by meta_fixed() and meta_random()
* Effect estimates of fitted meta-analysis models can be averaged by bma()
* Inclusion Bayes factor are computed by inclusion()
* User-specified and default prior functions are specified via prior() [can be plottet via plot(prior)]