#12 closed defect (fixed)
The Estimate function gives back a wrong standard deviation
Reported by: | imendez | Owned by: | Víctor de Buen Remiro |
---|---|---|---|
Priority: | high | Milestone: | |
Component: | Math | Version: | head |
Severity: | normal | Keywords: | |
Cc: |
Description
The Estimate function returns the norm instead of the standard deviation of the
residuals.
I have assigned the "TYPE_Set" component because the option "Estimate" does not
exist.
Change History (7)
comment:1 Changed 21 years ago by
Owner: | changed from danirus to jsperez |
---|
comment:2 Changed 21 years ago by
Component: | TYPE_Set → ESTIMATE |
---|
Moved to the new component "ESTIMATE".
Sorry for the inconvenience, imendez!
comment:3 Changed 21 years ago by
Owner: | changed from jsperez to Víctor de Buen Remiro |
---|
reassigned to Victor
comment:4 Changed 21 years ago by
Status: | new → assigned |
---|
comment:5 Changed 21 years ago by
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
In older versions of TOL the norm to be minimized was just the residuals
Frobenius norm. However this norm isn't now the same thing now.
So, there was an error in the reporting of the standard error at the end of
estimation.
At line 132 of file tol/btol/bmodel/estim.cpp, the assignement
M_->standardError_ = Sqrt(MtMSqr(M_->A_ )(0,0)/M_->A_.Rows());
solves the error.
comment:6 Changed 18 years ago by
bug_file_loc: | → http://cvs.tol-project.org/viewcvs.cgi/tol_tests/tol/ |
---|
comment:7 Changed 18 years ago by
bug_file_loc: | http://cvs.tol-project.org/viewcvs.cgi/tol_tests/tol/ → http://cvs.tol-project.org/viewcvs.cgi/tol_tests/tol/ARIMA/FunEstimate/ |
---|
Estimation Engine Bug Report