Fix dispatch and update documentation

[HenriDeh]

See #897 for the explanation.
I don't know how this affects current implementations of algorithms but it is likely to fix some rather than break anything.

PR Checklist

• Update NEWS.md?
• Unit tests for all structs / functions?
• Integration and correctness tests using a simple env?
• PR Review?
• Add or update documentation?
• Write docstrings for new methods?

[codecov]

Codecov Report

Merging #907 (2fa4b2d) into main (62e6e38) will increase coverage by 0.00%.
The diff coverage is 4.54%.

@@           Coverage Diff           @@
##             main     #907   +/-   ##
=======================================
  Coverage   26.01%   26.01%           
=======================================
  Files         220      220           
  Lines        7719     7726    +7     
=======================================
+ Hits         2008     2010    +2     
- Misses       5711     5716    +5     

Impacted Files	Coverage Δ
src/ReinforcementLearningBase/src/interface.jl	57.47% <ø> (ø)
...rcementLearningCore/src/policies/q_based_policy.jl	0.00% <0.00%> (ø)
...ments/experiments/DQN/JuliaRL_BasicDQN_CartPole.jl	0.00% <ø> (ø)
...ts/experiments/DQN/JuliaRL_BasicDQN_MountainCar.jl	0.00% <ø> (ø)
...periments/DQN/JuliaRL_BasicDQN_PendulumDiscrete.jl	0.00% <ø> (ø)
...ments/DQN/JuliaRL_BasicDQN_SingleRoomUndirected.jl	0.00% <ø> (ø)
...xperiments/experiments/DQN/JuliaRL_DQN_CartPole.jl	0.00% <ø> (ø)
...riments/experiments/DQN/JuliaRL_DQN_MountainCar.jl	0.00% <ø> (ø)
...xperiments/experiments/DQN/JuliaRL_IQN_CartPole.jl	0.00% <ø> (ø)
...riments/experiments/Offline/JuliaRL_BC_CartPole.jl	0.00% <0.00%> (ø)
... and 12 more

... and 3 files with indirect coverage changes

[HenriDeh]

A consequence of this is that any algorithm must manually implement the for batch in trajectory loop as we cannot know at which stage the loop must live.

[HenriDeh]

Tests should pass now. I will make some edits to the documentation before merging.