ERROR Malformed floating point number
Closed this issue · 13 comments
I have an error that does not allow the simulation to continue, the message is:
Error: Problem with keyword TIGHTEN
Internal error: Malformed floating point number 'RPTRST'
Error: Unrecoverable errors were encountered while loading input
Can you help me by telling what went wrong please
Thanks
Please post a section of deck where the error is.
Error: Problem with keyword TIGHTEN
In /home/ubuntu/Escritorio/OPM/OPM Data/EN_2019.sim/HM_FP_POES_2020-09-02/HM_FP_POES_2020-09-02_SCH.INC line 16.
Internal error: Malformed floating point number 'SKIPREST'
Error: Unrecoverable errors were encountered while loading input
the error is generated when reading data from the different .include files
I need to see the actual section in the input file may be 10 lines above and below in order to help.
sorry, from the .include file?
Yes the input were the error occurs in the include file +/- 10 lines.
Error: Problem with keyword TIGHTEN .../HM_FP_POES_2020-09-02_SCH.INC line 16. Internal error: Malformed floating point number 'SKIPREST'
I need to see the actual section in the input file may be 10 lines above and below in order to help.
As @OPMUSER says, it's generally not feasible to diagnose problems like these without having access to at least a small section of the input—in this case the HM_FP_POES_2020-09-02_SCH.INC
include file—near where error occurs. That said, and because I've seen similar symptoms ("malformed floating point number...") before, I'm going to guess that you've not properly terminated the TIGHTEN
keyword.
This keyword needs to be entered as
TIGHTEN
1.23 /
with a terminating /
character following the single floating-point value, either on the same line or on the subsequent line.
Addendum
I suspect that your include file has text that looks like
TIGHTEN
0.123
SKIPREST
If so, you need to insert the terminating slash character e.g.,
TIGHTEN
0.123 /
SKIPREST
RPTONLY
TUNING
1 0.5 1 0.15 3 0.3 0.4 1.5 1E20 /
0.1 0.001 5E-7 0.005 10 5 5E-6 0.05 0.001 0.025/
12 1 99 1 15 8 1E6 1E6 1E6 1E6/
TIGHTEN
1 /
SKIPREST
RPTSCHED
FIP=2 WELLS /
RPTRST
BASIC=3 FLOWS FREQ /
SKIP
ENDSKIP
GRUPTREE
'GROUP 1' FIELD /
/
DATES
29 JUL 2003 /
/
thanks for the help, this is the input file.
This should work, hopefully:
RPTONLY
TUNING
1 0.5 1 0.15 3 0.3 0.4 1.5 1E20 /
/ NOT SUPPORTED
/ NOT SUPPORTED
--
-- NOT SUPPORTED
--
-- TIGHTEN
-- 1 /
SKIPREST
RPTSCHED
FIP=2 WELLS=2 /
RPTRST
BASIC=3 FREQ=1 /
--
-- THIS HAS NO EFFECT BECAUSE THERE IS NOTHING IN BETWEEN SO COMMENT OUT
-- SKIP
-- ENDSKIP
GRUPTREE
'GROUP 1' FIELD /
/
DATES
29 JUL 2003 /
/
A few things to note:
- Only the first line of the
TUNING
keyword is used if, and only if, you setenable-tuning=true
in the parameter file; so I commented out the other two lines to avoid confusion. - The
TIGHTEN
keyword is not supported, so I commented it out to avoid confusion. - For
RPTSCHED
, I suspect that the WELLS sub-keyword is the issue, as it should be of the same form as the FIP sub-keyword, that isWELLS=2
. - Similarly for
RPTRST
and theFREQ
sub-keyword and FLOWS sub-keyword is not supported according to the manual.
Try and be more neat in your input decks as it helps in catching errors and makes the deck more readable. If you download OPMRUN (https://github.com/OPM/opm-utilities/tree/master/opmrun) then there is a keyword editor that will generate the keywords for you in a similar format as the examples in the manual.
ok, thanks a lot
@rcgz: Are you satisfied that the problem has been resolved? If so, please let us know or close the Issue yourself. If not, please provide additional details of what problems you see?
that problem is no longer there, but now when starting the report step it only reaches the middle and then generates an error:
<<Error: Unrecoverable errors were encountered while loading input>>
when starting the report step it only reaches the middle and then generates an error:
Error: Unrecoverable errors were encountered while loading input
I see. Thanks for the update. In this case, I believe we'll need to see at least the full .DBG
file from your failed simulation run, but it's likely that we'll need to see the actual input files too. The above diagnostic is emitted when the input reader encounters something that it cannot make sense of. It might be an unterminated keyword or record or it might be an unexpected number or type of items in an individual keyword.
I recently amended the diagnostic message (PR OPM/opm-simulators#3912) to include additional information that might aid the user in identifying the problem, but unless you'd like to build the simulator from source code yourself that facility won't be available to you until the 2022.10 release later this year.