| Version | Change log |
| gretl Portable 2023b Jul 21, 2023 |
- New command "kdplot": simple access to a kernel density plot for a single series - New function typename(): return a string giving the type of the argument (can replace typeof() and typestr()) - "arima" command: implement built-in procedure for assessing lag orders, via new --lagselect option, and add GUI access - "corr" command: add 1 percent critical value of correlation coefficient to the output - "join" command: import descriptions of newly-added series, if present - "gnuplot" command: add new options --inbuf and --outbuf to accept, respectively, input from a named string and output to a named string - "smpl" command with the panel-data --time option: support the --replace modifier - "outfile" command: insist on correct syntax for usage with the --tempfile option - aggregate() function: support matrix input - misszero() and zeromiss() functions: support matrix input - SVAR addon: implement mixed (set-based and zero) restrictions |
| gretl Portable 2021b May 5, 2021 |
New function trigamma(): second derivative of lngamma New function midasmult(): gives MIDAS multipliers New command "bds": BDS nonlinearity test Gdtb data format: make the new variant the default Tighten up on defbundle() syntax: no empty arguments allowed Array definition: enforce the specific choice of type "wls" command, with 0/1 weights: compute $yhat and $uhat for the excluded observations "dpanel" estimation via the GUI: make the --dpdstyle flag configurable "dpanel" --verbose switch: provide some information on the instruments used "adf" command: support more accurate inference with improved p-values or critical values, for the GLS case in particular "xtab" command: clarify that the $result (matrix) accessor is just for the bivariate case "clear" command: add --functions option to remove all hansl functions from memory "restrict" command: add "inject" keyword to support use of an array of strings to specify restrictions |
| gretl Portable 2021a Apr 14, 2021 |
- New function vma() for multiple time series - quantile(): support variant methods Q7 and Q8 described in Hyndman and Fan (1996) - defbundle(): add two shorthand variants of this function - irf(): support calculation of multiple impulse responses in a single call (with internal speed-up) - irf() bug-fix: failing to compute bootstrap confidence band correctly when passed a $system bundle argument - VAR internals: scrap augmented Cholesky matrix; so the $system.C accessor is now a square matrix - $system bundle: ensure presence of xlist member, and include the command-word (var, vecm or system) - mread(): support reading gdt and gdtb files as matrices - readfile(): support reading gzipped files transparently - obslabel(): support a vector of observations - nls/mle/gmm blocks: support use of printf statements - "open" command: support reading selected series from native gretl datafiles (gdt, gdtb) - "join" command: support $obsmajor, $obsminor as outer keys |
| gretl Portable 2020a B2020- Mar 6, 2020 |
Verbose output for numerical optimization: make the default more basic but add a new "full" setting for the set-variable "max_verbose" "wls" command: support the --cluster option "scatters" command: improve handling of daily data Work on lasso back-end (not public yet) |
| gretl Portable 2019d B2019- Dec 23, 2019 |
Fix bug: "dataset sortby" not handling missing values correctly on MS Windows Fix bug: possible crash on using the "fcast" command for out-of-sample forecasting after estimation of a VAR Fix bug: typeof() function not working correctly for string objects inside bundles and arrays Fix bug: robust variant of "chow" failing in case of exact collinearity in the augmented regression |
| gretl Portable 2019c B2019- Jul 3, 2019 |
Fix bug: "dataset sortby" not handling missing values correctly on MS Windows Fix bug: possible crash on using the "fcast" command for out-of-sample forecasting after estimation of a VAR Fix bug: typeof() function not working correctly for string objects inside bundles and arrays Fix bug: robust variant of "chow" failing in case of exact collinearity in the augmented regression |
