Christian gave me a feedback about the tutorial of scipp
that the setting variances in the DataArray propagates all the way to the original Variable.

It was
https://scipp.github.io/getting-started/quick-start.html
This page,

array = array.copy()
array.variances = np.square(rng.random((4, 5)))
sc.show(array)

He tried deleting array = array.copy() in this part and found it out.

I agree that it can be surprising to users.
Should we add the explanation whenever we have copy in the tutorial?

Or can we make it in a way that setting variances on DataArray doesn't affect the original Variable that became the DataArray.data?