Bug #1768 fixed authored by Sébastien Villemot's avatar Sébastien Villemot
...@@ -16,7 +16,7 @@ This page documents known bugs in Dynare stable. ...@@ -16,7 +16,7 @@ This page documents known bugs in Dynare stable.
* Output of `smoother`-results when using particle filters will be based on `order=1` (bug fixed, in 417f038c) * Output of `smoother`-results when using particle filters will be based on `order=1` (bug fixed, in 417f038c)
* Output of `moments_varendo`-results when using particle filters will be based on `order=1` (bug fixed in dedfd0c0, 314948ce. and b38d6f81) * Output of `moments_varendo`-results when using particle filters will be based on `order=1` (bug fixed in dedfd0c0, 314948ce. and b38d6f81)
* When decreasing the `order` in mod-files, `oo_.dr` may contain stale results from higher orders (bug fixed in 827a7fca) * When decreasing the `order` in mod-files, `oo_.dr` may contain stale results from higher orders (bug fixed in 827a7fca)
* Estimation results using the particle filter at `order=3` will be incorrect if the restricted state space differs from the unrestricted one (bug not yet fixed, see https://git.dynare.org/Dynare/dynare/-/issues/1768) * Estimation results using the particle filter at `order=3` will be incorrect if the restricted state space differs from the unrestricted one (bug fixed, see #1768)
* `mode_compute=102` (SOLVEOPT) may return with `Inf` instead of the last feasible value (bug fixed in a89f21b5) * `mode_compute=102` (SOLVEOPT) may return with `Inf` instead of the last feasible value (bug fixed in a89f21b5)
* Using `analytic_derivation` for Bayesian estimation will result in wrong results when the multivariate Kalman filter enters the steady state stage (bug fixed in c6c9b4e3) * Using `analytic_derivation` for Bayesian estimation will result in wrong results when the multivariate Kalman filter enters the steady state stage (bug fixed in c6c9b4e3)
* Using `analytic_derivation` for ML estimation will result in a crash (bug fixed in 82aa669b) * Using `analytic_derivation` for ML estimation will result in a crash (bug fixed in 82aa669b)
... ...
......