Create a function to export the whole solution to file (e.g., CSV, parquet, JSON, etc.)

[abelsiqueira]

Currently, we only export the assets investment.

[abelsiqueira]

Related to #115

[abelsiqueira]

How should we write the flow output? It is indexed in (u,v), rp, and B, where B is the time block.

• Do we write the whole time block variable as a column?
• Or do we divide the flow value between all time steps?

I.e., if flow[("a","b"), 1, 1:3] = 30.0, should we write

from,to,rp,B,value
a,b,1,1:3,30.0

or

from,to,rp,time_step,value
a,b,1,1,30.0
a,b,1,2,30.0
a,b,1,3,30.0

[clizbe]

Personally I prefer the second, but I guess it depends how you want to analyse the data.
@gnawin What do you think?

[gnawin]

@clizbe For me as well the second is more useful, because it makes it easier to manipulate the results further.

I have another conceptual question though, if flow[("a","b"), 1, 1:3] = 30, does it mean flow[("a","b"), 1, 1] = flow[("a","b"), 1, 2] = flow[("a","b"), 1, 3] = 30 or flow[("a","b"), 1, 1] = flow[("a","b"), 1, 2] = flow[("a","b"), 1, 3] = 10?
Or more generally, when we define any variable over a time block, is the value the sum over the time block, or of every time step in the time block? @datejada @abelsiqueira

[clizbe]

For storage it's energy, for flows it's power. We should add this explanation of exactly what the output means to the documentation.

[datejada]

With @suvayu, we discussed the new duckDB workflow, which will cover this part as well as exporting the outputs to other types of files. So, we added it to the epic issue #289

[abelsiqueira]

I will create a simple version to be used right now, independent of DuckDB.