1: \begin{abstract}

2: Subspace recycling iterative methods and other subspace augmentation schemes are a successful extension to Krylov subspace methods in which

3: a Krylov subspace is augmented with a fixed subspace spanned by vectors deemed to be helpful in accelerating convergence or conveying knowledge

4: of the solution.  Recently, a survey was published, in which a framework describing the vast majority of such methods was proposed 

5: [Soodhalter et al, GAMM-Mitt. 2020].  In many of these methods, the Krylov subspace is one generated by the system matrix composed with a projector

6: that depends on the augmentation space.  However, it is not a requirement that a projected Krylov subspace be used.  There are augmentation

7: methods built on using Krylov subspaces generated by the original system matrix, and these methods also fit into the general framework.  

8: 

9: In this note, we observe that one gains implementation benefits by considering such augmentation methods with unprojected Krylov subspaces

10: in the general framework.  We demonstrate this 

11: by applying the idea to the R$^3$GMRES method proposed in [Dong et al. ETNA 2014] to obtain a simplified implementation and to connect that algorithm

12: to early augmentation schemes based on flexible preconditioning [Saad. SIMAX 1997].

13: \end{abstract}

14: