hBayesDM - hBayesDM 1.3.1

• Add plot functions for Hierarchical Gaussian Filter models: plot_hgf_ibrb, plot_hgf_ibrb_single.

- TeX
Published by bugoverdose 6 months ago

hBayesDM - hBayesDM 1.3.0

• Added a Hierarchical Gaussian Filter model for binary inputs and binary responses: hgf_ibrb for hierarchical Bayesian analysis and hgf_ibrb_single for individual Bayesian analysis.
• Added new article: Hierarchical Bayesian Analysis on Hierarchical Gaussian Filter

- TeX
Published by bugoverdose 6 months ago

hBayesDM - hBayesDM 1.2.1

Fixed a pkgdown error.

- TeX
Published by hydoh over 3 years ago

hBayesDM - hBayesDM 1.2.0

• Added a drift diffusion model and two reinforcement learning-drift diffision models for the probabilistic selection task: pstRT_ddm, pstRT_rlddm1, and pstRT_rlddm6.
• Added multiple models for the banditNarm task: banditNarm_2par_lapse, banditNarm_4par, banditNarm_delta, banditNarm_kalman_filter, banditNarm_lapse, banditNarm_lapse_decay, and banditNarm_singleA_lapse.
• Fixed bart_ewmv to avoid dividing by zero.

- TeX
Published by hydoh over 3 years ago

hBayesDM - hBayesDM 1.1.1

• Fix the symlink error in the Python version due to releasing with poetry
• Fix minor errors in both R and Python

- TeX
Published by JaeyeongYang almost 5 years ago

hBayesDM - hBayesDM 1.1.0

• Added the cumulative model for the Cambridge gambling task: cgt_cm.
• Added two new models for aversive learning tasks: alt_delta and alt_gamma.
• Added exponential-weight mean-variance model for BART task: bart_ewmv.
• Added simple Q learning model for the probabilistic selection task: prl_Q.
• Added signal detection theory model for 2-alternative forced choice task: task2AFC_sdt.

- TeX
Published by JaeyeongYang about 5 years ago

hBayesDM - hBayesDM 1.0.2

• Fix an error on using data.frame objects as data (#112).

- TeX
Published by JaeyeongYang over 6 years ago

hBayesDM - hBayesDM 1.0.1

• Minor fix on R and Python codes (R, #111).

- TeX
Published by JaeyeongYang over 6 years ago

hBayesDM - hBayesDM 1.0.0

Major changes

• Now, hBayesDM has both R and Python version, with same models included! You can run hBayesDM with a language you prefer!
• Models in hBayesDM are now specified as YAML files. Using the YAML files, R and Python codes are generated automatically. If you want to contribute hBayesDM by adding a model, what you have to do is just to write a Stan file and to specify its information! You can find how to do in the hBayesDM wiki (https://github.com/CCS-Lab/hBayesDM/wiki).
• Model functions try to use parameter estimates using variational Bayesian methods as its initial values for MCMC sampling by default (#96). If VB estimation fails, then it uses random values instead.
• The data argument for model functions can handle a data.frame object (#2, #98).
• choiceRT_lba and choiceRT_lba_single are temporarily removed since their codes are not suitable to the new package structure. We plan to re-add the models in future versions.
• The Cumulative Model for Cambridge Gambling Task is added (cgt_cm; #108).

Minor changes

• The tau parameter in all models for the risk aversion task is modified to be bounded to 0, 30.
• bart_4par is fixed to compute subject-wise log-likelihood (#82).
• extract_ic is fixed for its wrong rep function usage (#94, #100).
• The drift rate (delta parameter) in choiceRT_ddm and choiceRT_ddm_single is unbounded and now it is estimated between -Inf, Inf.
• Fix a preprocessing error in choiceRT_ddm and choiceRT_ddm_single (#95, #109).
• Fix igt_orl for a wrong Matt trick operation (#110).

- TeX
Published by JaeyeongYang over 6 years ago

hBayesDM - hBayesDM 0.7.2

• Add three new models for the bandit4arm task: bandit4arm_2par_lapse, bandit4arm_lapse_decay and bandit4arm_singleA_lapse.
• Fix various (minor) errors.

- TeX
Published by JaeyeongYang about 7 years ago

hBayesDM - hBayesDM 0.7.1

• Make it usable without manually loading rstan.
• Remove an annoying warning about using ..insensitive_data_columns.

- TeX
Published by JaeyeongYang about 7 years ago

hBayesDM - hBayesDM 0.7.0

• Now, in default, you should build a Stan file into a binary for the first time to use it. To build all the models on installation, you should set an environmental variable BUILD_ALL to true before installation.
• Now all the implemented models are refactored using hBayesDM_model function. You don't have to change anything to use them, but developers can easily implement new model now!
• We added a Kalman filter model for 4-armed bandit task (bandit4arm2_kalman_filter; Daw et al., 2006) and a probability weighting function for general description-based tasks (dbdm_prob_weight; Erev et al., 2010; Hertwig et al., 2004; Jessup et al., 2008).
• Initial values of parameter estimation for some models are updated as plausible values, and the parameter boundaries of several models are fixed (see more on issue #63 and #64 in Github).
• Exponential and linear models for choice under risk and ambiguity task now have four model regressors: sv, sv_fix, sv_var, and p_var.
• Fix the Travix CI settings and related codes to be properly passed.

- TeX
Published by JaeyeongYang about 7 years ago

hBayesDM - hBayesDM 0.6.3

• Update the dependencies on rstan (>= 2.18.1)
• Remove rstantools from Imports
• No changes on model files, as same as the version 0.6.2

- TeX
Published by JaeyeongYang over 7 years ago

hBayesDM - hBayesDM 0.3.0

• Made several changes following the guidelines for R packages providing interfaces to Stan.
• Stan models are precompiled and models will run immediately when called.
• The default number of chains is set to 4.
• The default value of adapt_delta is set to 0.95 to reduce the potential for divergences.
• The printFit function uses LOOIC by default. Users can select WAIC or both (LOOIC & WAIC) if needed.

- TeX
Published by youngahn over 7 years ago

hBayesDM - hBayesDM 0.3.1

• Remove dependence on the modeest package. Now use a built-in function to estimate the mode of a posterior distribution.
• Rewrite the "printFit" function.

- TeX
Published by youngahn over 7 years ago

hBayesDM - hBayesDM 0.4.0

• Add a choice reaction time task and evidence accumulation models
  • Drift diffusion model (both hierarchical and single-subject)
  • Linear Ballistic Accumulator (LBA) model (both hierarchical and single-subject)
• Add PRL models that can fit multiple blocks
• Add single-subject versions for the delay discounting task (dd_hyperbolic_single and dd_cs_single).
• Standardize variable names across all models (e.g., rewlos --> outcome for all models)
• Separate versions for CRAN and GitHub. All models/features are identical but the GitHub version contains precompilled models.

- TeX
Published by youngahn over 7 years ago

hBayesDM - hBayesDM 0.6.2

What's new in v0.6.2

• Fix an error on choiceRT_ddm (#44)

- TeX
Published by JaeyeongYang over 7 years ago

hBayesDM - hBayesDM 0.6.1

What's new in v0.6.1

• Solve an issue with built binary files.
• Fix an error on peer_ocu with misplaced parentheses.

- TeX
Published by JaeyeongYang over 7 years ago

hBayesDM - hBayesDM 0.6.0

What's new in v0.6.0

• Add new tasks (Balloon Analogue Risk Task, Choice under Risk and Ambiguity Task, Probabilistic Selection Task, Risky Decision Task (a.k.a. Happiness task), Wisconsin Card Sorting Task)
• Add a new model for the Iowa Gambling Task (igt_orl)
• Change priors (Half-Cauchy(0, 5) --> Half-Cauchy(0, 1) or Half-Normal(0, 0.2)
• printFit function now provides LOOIC weights and/or WAIC weights

- TeX
Published by youngahn over 7 years ago

hBayesDM - hBayesDM 0.5.1

What's new in v0.5.1.

• Add models for the Two Step task
• Add models without indecision point parameter (alpha) for the PRL task (prl*woa.stan)
• Model-based regressors for the PRL task are now available
• For the PRL task & prlfictitious.stan & prlfictitious_rp.stan --> change the range of alpha (indecision point) from [0, 1] to [-Inf, Inf]

- TeX
Published by youngahn almost 8 years ago

hBayesDM - hBayesDM 0.5.0

What's new in v0.5.0

• Support variational Bayesian methods (vb=TRUE)
• Allow posterior predictive checks, except for drift-diffusion models (inc_postpred=TRUE)
• Add the peer influence task (Chung et al., 2015, USE WITH CAUTION for now and PLEASE GIVE US FEEDBACK!)
• Add 'prlfictitiousrp' model
• Made changes to be compatible with the newest Stan version (e.g., // instead of # for commenting).
• In 'prl_*' models, 'rewlos' is replaced by 'outcome' so that column names and labels would be consistent across tasks as much as possible.
• Email feature is disabled as R mail package does not allow users to send anonymous emails anymore.
• When outputs are saved as a file (*.RData), the file name now contains the name of the data file.

- TeX
Published by youngahn about 8 years ago