1: \begin{abstract}

2: According to Haar's Theorem, every compact topological group $G$

3: admits a unique (regular, right and) left-invariant Borel probability measure $\mu_G$.

4: Let the \emph{Haar integral} (of $G$) denote the functional

5: $\int_G:\calC(G)\ni f\mapsto \int f\,d\mu_G$

6: integrating any continuous function $f:G\to\IR$ with respect to $\mu_G$.

7: This generalizes, and recovers

8: for the additive group $G=[0;1)\mod 1$,

9: %~aka multiplicative complex circle group $\calU(1)\cong\calSO(2)$,

10: the usual Riemann integral: computable (cmp. Weihrauch 2000, Theorem 6.4.1),

11: and of computational cost characterizing complexity class \classSharpP$_1$

12: (cmp. Ko 1991, Theorem 5.32).

13: 

14: We establish that in fact, every computably compact computable metric group

15: renders the Haar measure/integral computable: 

16: once asserting computability using an elegant synthetic argument,

17: exploiting uniqueness in a computably compact space of probability measures;

18: and once presenting and analyzing an explicit, imperative algorithm

19: based on `maximum packings' with rigorous error bounds and guaranteed convergence.

20: Regarding computational complexity, for the groups $\calSO(3)$ and $\calSU(2)$,

21: we reduce the Haar integral to and from Euclidean/Riemann integration.

22: In particular both also characterize \classSharpP$_1$.

23: Implementation and empirical evaluation using the iRRAM C++ library for exact real computation

24: confirms the (thus necessary) exponential runtime.

25: \end{abstract}

26: