I'm trying to patch armax. The key moment is treating of $error

Command ref. says one should use catch prefix alongside.

I tried context search for "catch" in armax.gfn and didn't find any.

Is $error behavior different in loops?

A fragment for quick reference

# begin of armax.gfn fragment 

loop q = 0..maxq -q

      # estimate model

      if nelem(xlist)=0

        if p=0 && q=0

          # p,q=0 時, 一定要有 const, 不然會有參數數目不足之 error

          arima p q ; y -q @eMethod

        else

          arima p q ; y -q @eMethod @sCons

        endif

      else

        arima p q ; y xlist -q @eMethod @sCons

      endif

      err = $error

      if (err)

        printf "Got error %d (%s)\n", err, errmsg(err)

      endif

      if isnull($aic)=0

        printf "Got error,p=%d,q=%d\n",p,q

      else

        @tabname[nrow,1] = p

        @tabname[nrow,2] = q

        @tabname[nrow,3] = $aic

        @tabname[nrow,4] = $bic

        @tabname[nrow,5] = $hqc

        # test for mininum noise model

        if nrow=1

          min_aic = $aic

          min_sc = $bic

          min_hq = $hqc

          order_aic = nrow

          order_sc = nrow

          order_hq = nrow

        else

          if $aic < min_aic

            min_aic = $aic

            order_aic = nrow

          endif

          if $bic < min_sc

            min_sc = $bic

            order_sc = nrow

          endif

          if $hqc < min_hq

            min_hq = $hqc

            order_hq = nrow

          endif

        endif

      endif

      nrow = nrow+1

    endloop

# end of armax fragment