py and libm: Add asinf,acosf; print higher precision for float.

Also use less stack space when printing single precision float.

Addition of asinf and acosf addresses issue #851.