question: expected behavior of conventional vs primitive cell?

[bkmi]

If a (conventional) cell contains multiple exact copies of the primitive cell, the algorithm produces a different hash.

Is this expected / desired behavior?

It obviously can be averted by first converting both systems to primitive cells; however, with disorder this can be an issue.

Ben

[msiron-entalpic]

Hi @bkmi you're absolutely right that ideally primitive cells should return the same hash as the standard cell, as should supercell and other kinds of transformation. I made a PR #21 , would you mind taking a look? It is in branch fix/bawl-primitive.

I am unsure its fair to compare BAWL with a primitive structure to methods where such transformation has not been applied.