impl Mul<Array<Complex<T>>> for T

[cavemanloverboy]

I did a quick scan of the issues and didn't see this. For an array: Array<Complex<f64>> (for example), i think it's perfectly reasonable to expect

let output: Array<Complex<f64>> = 2.0_f64 * array;

to work. I've only ever gotten a rhs mul e.g. array * 2.0 to work. (see example, non-working example). Is there a reason this hasn't been done, e.g. orphan rules?

[cavemanloverboy]

I think it should be possible, actually, here's an example

[cavemanloverboy]

The implementation can also be done for Complex<f32> and Complex<f64> for Array<Complex<f32>> and Array<Complex<f64>>

[cavemanloverboy]

Also, doing this for Div will allow for things like 1.0 / array for a complex array.

[bluss]

Please see this for doc on scalars on the left hand side: https://docs.rs/ndarray/latest/ndarray/trait.ScalarOperand.html

Just for baseline/background. I don't know if this is a good idea, don't know if there are any drawbacks. Since complex implements mixed float-complex operations, we have the necessary precedent to consider it.

[akern40]

I believe this would be closed by #782? The next-gen compiler will enable that PR and seems to be close to stabilization, as it is a Rust goal for 2024.