Combine multiple DSS files into one for post-processing, and model comparison.
python -m blendss "studies.toml" --fv "calsim3"
I have two DSS files, as described below:
- DSS #1, This is a DSS file that contains model outputs from a published study. One of the outputs is
/MODEL/RESULT/FLOW///L2020A/
. - DSS #2, This is a DSS file that I made, which contins model outputs from my own study. The
/MODEL/RESULT/FLOW///L2020A/
is also in my dataset.
I want to compare the results of these models by looking at /MODEL/RESULT/FLOW///L2020A/
. The problem is, I need to carry around both of these DSS files if I want to do that.
I need to blend these DSS files together!
If I use blendss
, I can combine the two DSS files like this:
- From DSS #1, we copy
/MODEL/RESULT/FLOW///L2020A/
to the new DSS file, renaming it to/PUBLISED/RESULT/FLOW///L2020A/
. - From DSS #2, we copy
/MODEL/RESULT/FLOW///L2020A/
to the new DSS file, renaming it to/MINE/RESULT/FLOW///L2020A/
.
Since we a are changing the A-part
of the DSS path, these two datasets don't collide. We can choose the new A-part
to be descriptive, so the content semi-self-docuemnts.
How do I do this with blendss
?
- Create a
toml
configuration file that looks like this:
[[study]]
a="PUBLISHED"
dss="/path/to/DSS#1.dss"
[[study]]
a="MINE"
dss="/path/to/DSS#2.dss"
- Create a
fv
file that looks like this:
B | C |
---|---|
RESULT | FLOW |
- Run blendss:
python -m blendss "path/to/your/config.toml" --fv "/path/to/your/config.fv"
You can even use one of our built-in fv files like this:
python -m blendss "path/to/your/config.toml" --fv "calsim3"
calsim3.fv
: Standard paths read from the CalSim3 DV file.
flowchart LR
BLENDER[[blendss]]
BLEND[(`blended.dss`)]
A[(base_study.dss)]
B[(exploration_study.dss)]
C[(published_study.dss)]
A --> BLENDER
B --> BLENDER
C --> BLENDER
BLENDER --> BLEND