Assertion failed for localGridIndex

[opm]

Assertion failed for localGridIndex

[JacobStoren]

This seems to be due to some inconsistency in the data ERT provides.
When we do :

int numLGRs = ecl_grid_get_num_lgr(mainEclGrid);
int localCellIdx = 639;
const ecl_grid_type* subGrid = ecl_grid_get_cell_lgr1(mainEclGrid, localCellIdx);
int subGridFileIndex = ecl_grid_get_lgr_nr(subGrid );

the subGridFileIndex ends up as 110, while numLGRs is only 109. I think the ERT team needs to look into this.

[joakim-hove]

Hello,

I think the behaviour is right, but it has changed and I'm reallly sorry
you have not got info.

The point is that for each LGR the EGrid files have a header which contains
a number Id for this lgr (lgr_nr). In 99% of the cases this nr coresponds
running number of the lgr in the file, but for the particular file you are
considering there is actually a hole in the series, lgr_nr 108 is missing.

So I think this must be fixed in Resinsight, but sorry for the poor
communication.

Joakim
Den 5. sep. 2013 00:02 skrev "Jacob Støren" [email protected]
følgende:

This seems to be due to some inconsistency in the data ERT provides.
When we do :

int numLGRs = ecl_grid_get_num_lgr(mainEclGrid);int localCellIdx = 639;const ecl_grid_type* subGrid = ecl_grid_get_cell_lgr1(mainEclGrid, localCellIdx);int subGridFileIndex = ecl_grid_get_lgr_nr(subGrid );

the subGridFileIndex ends up as 110, while numLGRs is only 109. I think
the ERT team needs to look into this.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/43#issuecomment-23823696
.

[joakim-hove]

That was not really clear from my first e-mail: Previously the lgr_nr
corresponded directly to the running number in the file.

Now lgr_nr is actually read from the header and in some very few cases
(this being the only I have seen) they disagree.

Joakim
Den 5. sep. 2013 08:16 skrev "Joakim Hove" [email protected] følgende:

Hello,

I think the behaviour is right, but it has changed and I'm reallly sorry
you have not got info.

The point is that for each LGR the EGrid files have a header which
contains a number Id for this lgr (lgr_nr). In 99% of the cases this nr
coresponds running number of the lgr in the file, but for the particular
file you are considering there is actually a hole in the series, lgr_nr 108
is missing.

So I think this must be fixed in Resinsight, but sorry for the poor
communication.

Joakim
Den 5. sep. 2013 00:02 skrev "Jacob Støren" [email protected]
følgende:

This seems to be due to some inconsistency in the data ERT provides.
When we do :

int numLGRs = ecl_grid_get_num_lgr(mainEclGrid);int localCellIdx = 639;const ecl_grid_type* subGrid = ecl_grid_get_cell_lgr1(mainEclGrid, localCellIdx);int subGridFileIndex = ecl_grid_get_lgr_nr(subGrid );

the subGridFileIndex ends up as 110, while numLGRs is only 109. I think
the ERT team needs to look into this.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/43#issuecomment-23823696
.

[JacobStoren]

Ok, so the ecl_grid_get_lgr_nr( ) is actually retreiving an id, and not an index. Thanks

Jacob

[JacobStoren]

This is fixed now, so the model in question now loads.