Open jcrivenaes opened 2 months ago
Discussion 30 Sept 2024:
The consensus is that the context around this mapping needs to be matured and this is probably not a target for simplified functions yet. It may become one when approached from the top-down, cf #392.
Also that the format for .lyr
files, in general, seems odd and possibly unnecessary to support by itself. It might be better contained by a .yml
scheme with some script in fmu.tools
to translate it when it is needed or as metadata contained within the grid metadata.
Zone range information (zone->layer mapping) is used for several purposes. We are currently creating dataframes, .txt
, .csv
, and .lyr
files for layers/zonations in Drogon.
.yml
files are used for QC in GRID3D_HC_THICKNESS
and GRID3D_AVG_MAP
.csv
(verbose) zone->layer mapping (used in gendata_rft
?).txt
(verbose) zone->layer mapping. Readable by ResInsight.lyr
files can be used as input to
This information is needed for Eclipse grids as they dont have zonation information
SIM2SUMO
instead?The dataframe-generated format is uploaded to Sumo, none of the others are
One idea is to codify this data within grid metadata
fmu.tools
for these
.lyr
, .txt
, .csv
formats as needed.lyr
format:ResInsight has a page about this format here. Notably it also supports a color as a third column, which is not output by current Drogon exports and probably should not be.
It also supports the text format written below, calling is "FMU Layer Zone Table".
'Valysar' 1 - 10
'Therys' 11 - 27
'Volon' 28 - 32
.yml
format:zranges:
- Valysar: [1, 10]
- Therys: [11, 27]
- Volon: [28, 32]
.txt
format:1 Valysar
2 Valysar
...
11 Therys
12 Therys
...
28 Volon
29 Volon
...
.csv
/dataframe format:The dataframe exported by dataio has the column labels LAYER,ZONE
while "hand-made" csv exports do not.
1,Valysar
2,Valysar
...
11,Therys
12,Therys
...
28,Volon
29,Volon
...
The current Drogon example has now 6 script all doing zone to grid K counting mapping (different formats, different purpose). This becomes a burden of the end users, and there is a great potentiaal to simplify using a combinaion of
fmu.tools
andfmu.dataio
Propose to simplify according to the following style: