NEWS
fddm 2024-06-26
New Features
- added function
ddm() which permits formula-based fitting of the DDM using the analytic gradient.
Other Changes
- the order of the arguments
v and a to functions dfddm(), pfddm(), and the PDF derivatives d*_ddm() were changed. Previously, it was ...ddm(..., a, v, ...), and now it is ...ddm(..., v, a, ...)
- character variables in med_dec transformed to factors.
- updated example vignette by showing how to use the new
ddm() function.
fddm 2022-3-15
New Features
- Added new method to compute the PDF of the DDM
Bug Fixes
- Fixed HTML deprecation in documentation
Other Changes
- Changed the function parameters in
dfddm() that determine how the method of calculation is chosen.
fddm 2021-12-10
Bug Fixes
- Fixed 'pi' issue with
Rcpp
fddm 2021-7-27
New Features
- Added the function
pfddm() to calculate the cumulative distribution function (CDF) of the DDM; it uses the same model parameters as the existing dfddm() function.
Bug Fixes
- Fixed issue where the log optional parameter was ignored and the non-log result was produced.
fddm 2021-4-4
Bug Fixes
- Fixed issue on Solaris involving floor function and the int data type.
- Revamped handling of model parameters whose values are out-of-bounds.
fddm 2021-3-21
New Features
- Added new parameter
sigma to denote the diffusion coefficient of the Wiener process that underlies the DDM.
Bug Fixes
- Corrected the required number of terms in the small-time approximation methods.
k now consistently refers to the number of individual terms instead of sometimes the number of pairs of terms.
Other Changes
- Changed default value of
max_terms_large to 1. This affects the default method (n_terms_small = "SWSE", scale = "both").
- Revamped the error handling to make it less intrusive.
- Removed function call in
R; now directly calls C++.
- Updated vignettes.
fddm 2020-11-03
Other Changes
- Added warning about old 32-bit compilers exhibiting abnormal behavior for the large-time approximation.
fddm 2020-10-09
Bug Fixes
- The num_funcs.cpp functions are now prevented from possible overflow.
- Checks are now in place for empty strings as inputs to the parameters
n_terms_small, summation_small, and scale.
fddm 2020-10-07
New Features
- New default method, callable using the arguments
n_terms_small = "SWSE", scale = "both".
Bug Fixes
- Density functions now return strictly non-negative values.
Other Changes
- If input response times is an empty vector, then an empty vector is returned.
- Fine-tuned the summation functions used with the SWSE method
- Added URLs to DESCRIPTION
fddm 2020-07-10
Initial CRAN release.