for list comprehension of train_history

[BernhardAhrens]

Needed to make this [row.nse.SOC for row in o_SOC.val_history] work

Otherwise I would need to do this
[row.nse.SOC for row in o_SOC.val_history.data]

The .data is not very intuitive. Does not affect train (I tested), not sure if this would affect anything else @lazarusA

[lazarusA]

we need this.

The intended use of this is to first do something like soc = WrappedTuples(SOC.history), and then soc.nse to get all values, without the need to use list comprehension.

[BernhardAhrens]

EasyHybrid.jl/src/utils/wrap_tuples.jl

Line 8 in 841c7e6

struct WrappedTuples{T <: AbstractVector{<:NamedTuple}} <: AbstractVector{NamedTuple}

It's also just odd to have an 2D vector 😆

[BernhardAhrens]

What do we need it for? train and the train_board works. For saving?

[BernhardAhrens]

When I do soc.nse, it is still a bit weird format which is not straightforward to extract (for me ;-))

11×5 Matrix{@NamedTuple{SOC::Float32, CQ::Float32, CB::Float32, mean::Float32}}:
 (SOC = -16210.9, CQ = -27.1608, CB = -55.532, mean = -5431.19)   (SOC = -16210.9, CQ = -27.1608, CB = -55.532, mean = -5431.19)   …  (SOC = -16210.9, CQ = -27.1608, CB = -55.532, mean = -5431.19)
 (SOC = -86.3123, CQ = -23.1424, CB = -59.6994, mean = -56.3847)  (SOC = -86.3123, CQ = -23.1424, CB = -59.6994, mean = -56.3847)     (SOC = -86.3123, CQ = -23.1424, CB = -59.6994, mean = -56.3847)```

[lazarusA]

What do we need it for? train and the train_board works. For saving?

is for convenient access of nested arrays of tuples.

your output from above could further be nse_t = WrappedTuples(soc.nse) and then nse_t.SOC to get the eleven values of SOC.

[BernhardAhrens]

Hmm, I do not see the convenience of doing this
WrappedTuples(WrappedTuples(o_SOC.train_history).nse).SOC (Is that what you meant?)
over
[i.nse.SOC for i in o_SOC.val_history]

Both work with my small change btw.

This also related to Andrés issue I think #174

I think we need some convenience functions for this. Maybe we have them already and I forgot 😃