*************************************** * DYNARE TEST RESULTS * * for make check-matlab * *************************************** 142 tests PASSED out of 144 tests run *************************************** List of 2 tests FAILED: * gsa/ls2003.mod * conditional_forecasts/fs2000_est.mod ***************************************
*************************************** * DYNARE TEST RESULTS * * for make check-octave * *************************************** 136 tests PASSED out of 142 tests run *************************************** List of 6 tests FAILED: * fs2000/fs2000d.mod * gsa/ls2003.mod * identification/kim/kim2.mod * identification/as2007/as2007.mod * conditional_forecasts/fs2000_cal.mod * conditional_forecasts/fs2000_est.mod ***************************************
A full log can be found on karaba in /tmp/dynare-check-teosKunJLg.log
Hi there,
to wrap up the status of identification in 4.2:
for version 4.2.1, identification works for matlab >= 2009b for version 4.2.2, the same problem as 4.2.1 is present but on top of it I have introduced a new stupid bug trying to fix a tricky numerical problem in the check of the rank condition of Jacobians (I manually ported a similar fix from a commit in 4.3, but I did a mistake in doing that).
I have updated accordingly the wiki pages on known bugs. Apologies
Marco
-
Marco Ratto -
Sébastien Villemot