for i in 1:length(a)

a[i] += 1

end

I agree with your point, it's great to have all of this stuff out of the box and integrated with the language. Of course, though if you needed element-wise operations, you could go with some groovy dispatch

f(x) = [value * 22 / 7 + 5 for value in x]

and that will always produce a return as well, and won't mess with the data being passed through it. It's really could because you could write a normalization method that Sklearn needs to fit_transform with basically as you would assign a variable.

Julia is a great language.