1: \begin{abstract}  Let $(M,g)$ be a Riemannian manifold.  If $\mu$ is a probability measure on $M$

2: given by a continuous density function, one would expect the Fr\'{e}chet means of data-samples 

3: $Q=(q_1,q_2,\dots, q_N)\in M^N$, with respect to $\mu$, to behave ``generically''; 

4: e.g. the probability that the Fr\'{e}chet mean set $\FM(Q)$ has

5: any elements that lie in a given, positive-codimension submanifold, should be zero

6: for any $N\geq 1$. Even this simplest instance of genericity does not

7: seem to have been proven in the literature, except in special cases. 

8: The main result of this paper is a general, and stronger, genericity property:

9: given i.i.d. absolutely continuous $M$-valued random variables $X_1,\dots, X_N$,

10: and a subset 

11: $A\subset M$ of volume-measure zero, 

12: ${\rm Pr}\left\{\FM(\{X_1,\dots,X_N\})\subset M\backslash A\right\}=1.$

13: We also establish a companion theorem for {\em equivariant Fr\'{e}chet means},

14: defined when $(M,g)$ arises as  the quotient of a Riemannian manifold $(\tim,\tg)$ by a free,

15: isometric action of a finite group.  The equivariant Fr\'{e}chet means 

16: lie in $\tim$, but, as we show, project down to the ordinary Fr\'{e}chet sample means, 

17: and enjoy a similar genericity property.  Both these theorems are proven as consequences of a purely geometric 

18: (and quite general) result that constitutes the core mathematics in this paper:  

19: If $A\subset M$ has volume zero in $M$ , then the set $\{Q\in M^N : \FM(Q) \intersect A\neq\emptyset\}$ 

20: has volume zero in $M^N$.  We conclude the paper with an application 

21: to {\em partial scaling-rotation means}, a type of mean for symmetric positive-definite matrices.

22: 

23: 

24: \end{abstract}