1: \begin{abstract}

2:   A Dirichlet $k$-partition of a domain $U \subseteq \RR^d$ is a collection of

3:   $k$ pairwise disjoint open subsets such that the sum of their first

4:   Laplace-Dirichlet eigenvalues is minimal. A discrete version of Dirichlet

5:   partitions has been posed on graphs with applications in data analysis. Both

6:   versions admit variational formulations: solutions are characterized by

7:   minimizers of the Dirichlet energy of mappings from $U$ into a singular space

8:   $\Sigma_k \subseteq \RR^k$. In this paper, we extend results of N.\ Garc\'ia

9:   Trillos and D.\ Slep\v{c}ev to show that there exist solutions of the

10:   continuum problem arising as limits to solutions of a sequence of discrete

11:   problems. Specifically, a sequence of points $\{x_i\}_{i \in \NN}$ from $U$

12:   is sampled i.i.d.\ with respect to a given probability measure $\nu$ on $U$

13:   and for all $n \in \NN$, a geometric graph $\GGG_n$ is constructed from the

14:   first $n$ points $x_1, x_2, \ldots, x_n$ and the pairwise distances between

15:   the points. With probability one with respect to the choice of points

16:   $\{x_i\}_{i \in \NN}$, we show that as $n \to \infty$ the discrete Dirichlet

17:   energies for functions $\GGG_n \to \Sigma_k$ $\Gamma$-converge to (a scalar

18:   multiple of) the continuum Dirichlet energy for functions $U \to \Sigma_k$

19:   with respect to a metric coming from the theory of optimal transport. This,

20:   along with a compactness property for the aforementioned energies that we

21:   prove, implies the convergence of minimizers. When $\nu$ is the uniform

22:   distribution, our results also imply the statistical consistency statement

23:   that Dirichlet partitions of geometric graphs converge to partitions of the

24:   sampled space in the Hausdorff sense.

25: \end{abstract}

26: