I ran an optimization of a structure with stoichiometry YBaMn2O5F, and added a powerup to check the stability. The following was written to the "launches" section of my mongo database
"action": {
"stored_data": {
"task_id": 6,
"is_compatible": true,
"decomposes_to": [
{
"formula": "BaF2",
"amount": 0.15000000000000002,
"material_id": "MP#mp-1029"
},
{
"formula": "Ba5Mn5O14",
"amount": 0.05714285714285694,
"material_id": "MP#mp-763829"
},
{
"formula": "YMnO3",
"amount": 0.5,
"material_id": "MP#mp-19385"
},
{
"formula": "Mn3O4",
"amount": 0.1166666666666667,
"material_id": "MP#mp-18759"
},
{
"formula": "Ba8Mn8O21",
"amount": 0.17619047619047734,
"material_id": "MP#mp-694888"
}
],
"entry_id": null,
"e_above_hull": 0.10458794885714973
which would seem to give the stoichiometry
Y: 0.5
Ba: 1.85
Mn: 2.55
O: 6.47
F: 0.3
My second thought was that "amount" might mean "fraction of the cell in the referenced MP calculation". Since Mn3O4 has 3 formula units in the MP cell, YMnO3 has 6, Ba8Mn8O21 has 1, BaF2 has 1, and Ba5Mn5O14 has 2, this would give
Y: 3
Ba: 2.13
Mn: 6.03
O: 15.7
F: 0.3
This is much closer but a couple of the species are still off by a large amount. Could someone please let me know what the amounts mean?
They seem to be created by the Simplex class in pymatgen.util.coord, then passed through several functions in pymatgen.phasediagram.analyzer, then finally to check_stability in glue_tasks.py