FFTrees version 1.9.0 was released on CRAN [on 2023-02-08].
Apart from adding functionality and fixing minor bugs, this version improves consistency, robustness, and transparency.
Changes since last release:
Major changes
- Enabled optimizing for a user-defined
my.goal
on cue and tree levels (as defined bymy.goal.fun
). - Enabled optimizing for
dprime
on cue and tree levels (by using"dprime"
asgoal.threshold
,goal.chase
, orgoal
values). - Increased vocabulary for interpreting verbal FFT descriptions (using
my.tree
). - Improved
summary.FFTrees()
function:- Included current goal and cost values (if
"cost"
occurs in goals). - Included criterion base rates (in performance statistics on train and test data).
- Included current goal and cost values (if
Minor changes
- Included
dprime
values in cue level statistics (x$cues$thresholds
andx$cues$stats
). - Included
dprime
values in competition statistics (x$competition$train
andx$competition$test
). - Improved user feedback on combinations of goal and cost values.
- Prepared for modular tree translation and editing functions (
util_gfft.R
). - Prepared for global tree notation separator (
fft_node_sep
). - Added decision outcome and cue costs to
asif_results
(infftrees_grow_fan()
).
Details
- Added verification functions (for checking integrity of objects and validity of inputs).
- Deprecated the
rounding
argument ofFFTrees()
. - Re-arranged arguments of key functions (
FFTrees()
andfftrees_create()
) by functionality. - Re-arranged and cleaned code (in main and helper functions).
- Re-defined local constants as global constants (in
util_const.R
). - Revised status badges in
README
. - Tweaked plotting parameters.
- Fixed bugs and revised vignettes.
The current development version of FFTrees is available at https://github.com/ndphillips/FFTrees.