1: \begin{abstract}

2: In this work, we study a constrained monotone inclusion involving 

3: the 

4: normal cone to a closed vector subspace and a priori information

5: on primal solutions. We model this 

6: information by imposing that solutions belong to the fixed 

7: point set of an averaged nonexpansive mapping. We characterize the 

8: solutions using an auxiliary inclusion that involves the partial inverse 

9: operator. Then, we propose the 

10: primal-dual partial inverse splitting and we prove its weak 

11: convergence to a solution of the inclusion, generalizing several 

12: methods in the 

13: literature. The 

14: efficiency of the proposed method is illustrated in 

15: multiple applications including constrained LASSO,  

16: stochastic arc capacity expansion problems in transport 

17: networks, and variational mean field games with non-local couplings.

18: \par

19: \bigskip

20: 

21: \noindent \textbf{Keywords.} {\it Constrained convex optimization, 

22: Constrained 

23: LASSO, Monotone operator 

24: theory, Partial inverse method, Primal-dual splitting,

25:  Stochastic arc capacity expansion,

26: Mean field games.}

27: \end{abstract}

28: