The following functions currently return f64 values irrespective of the input Array data's native type.

• sum_all
• sum_nan_all
• product_all
• product_nan_all
• min_all
• max_all
• all_true_all
• any_true_all
• count_all
• imin_all
• imax_all

The above functions need to return the appropriate type value instead of f64. This change is probably going to be
easier to implemented once Typed Arrays are implemented.