need more specific description of feilds

[changxiaoju]

i begin to use dpti and it seems to work but questions below confuse me:
1)There is no FreeEnergyLiquid.json, can not open. And i don't find it in dpti pakages. Is it the same with FreeEnergy.json by settting "if_liquid": true ?
2)why are there three sets of lambda? what does lambda_soft_on , lambda_deep_on and lambda_soft_off mean?
3)why do hti.liquid.json need Einstein solid approximation?

[changxiaoju]

what do "if_meam" and "meam_model" mean?

[changxiaoju]

I have two elements so i set sigma_0_1, but it crushed with AssertionError: there must be key-value for sigma_0_0 in soft_param。so, what is the usage of sigma_0_0 and sigma_0_1,what do they mean?

[felix5572]

I will add more descriptions later.

1. Yes, I will update the json file soon.
2. This represents the integration path, lambda_soft_on: from pure einstein-solid to einstein-solid+LJ-potential,
   deep_on: from einstein-solid+LJ-potential to einstein-soilid+LJ-potential+deep-potential,
   soft-off: from einstein-soilid+LJ-potential+deep-potential, to only deep-potential
3. for liquid, we don't need Einstein approximation, we use ideal gas as reference.
   4

[felix5572]

if we set if_meam to true, we will use deep potential to replace deep potential.
meam_model: the mean model filenames

[felix5572]

in two pair LJ potential, the sigma values in LJ-potential's parameters can be different between elements.
sigma_0_0: element 0 and element 0 (defined by LAMMPS)
sigma_0_1: element 0 and element 1
sigma_1_1: element 1 and element 1

[changxiaoju]

I will add more descriptions later.

1. Yes, I will update the json file soon.
2. This represents the integration path, lambda_soft_on: from pure einstein-solid to einstein-solid+LJ-potential,
   deep_on: from einstein-solid+LJ-potential to einstein-soilid+LJ-potential+deep-potential,
   soft-off: from einstein-soilid+LJ-potential+deep-potential, to only deep-potential
3. for liquid, we don't need Einstein approximation, we use ideal gas as reference.
   4

1)ok,thanks
2)these three fields(lambda_soft_on , lambda_deep_on and lambda_soft_off) are in hti.liquid.json,which means they are for liquid thus using ideal gas as reference, so 2.1)why you mentined einstein-solid in explaination? And if i treat it as ideal gas , i still do not understand 2.2)what do you mean by '+'. like how you apply both lj and deep potential together? i think lambda only controls the weight of 'to' procedure.2.3) what should i do if i only have well trained deep potential?
3)yes,but i find "crystal": "frenkel", in hti.liquid.json, 3.1)what should i set it? and besides,3.2) what if i set "langevin": false,, what thermostat will be used then?

[changxiaoju]

if we set if_meam to true, we will use deep potential to replace deep potential. meam_model: the mean model filenames

in two pair LJ potential, the sigma values in LJ-potential's parameters can be different between elements. sigma_0_0: element 0 and element 0 (defined by LAMMPS) sigma_0_1: element 0 and element 1 sigma_1_1: element 1 and element 1

[changxiaoju]

if we set if_meam to true, we will use deep potential to replace deep potential. meam_model: the mean model filenames

what kind of situation to use it, replace a deep potential with another one.