ML_RCUT1

ML_RCUT1 = [real]
Default: ML_RCUT1 = 8.0 

Description: Sets the cutoff radius $R_{\text{cut}}$ for the radial descriptor ${\rho }_i^{(2)}(r)$ in $\AA$.

---

The radial descriptor for machine-learned force fields is constructed from

${\rho }_i^{(2)}\left(r\right)=\frac{1}{4\pi }\int {\rho }_i\left(r\widehat{\mathbf{𝐫}}\right)d\widehat{\mathbf{𝐫}},\text{where}{\rho }_i\left(\mathbf{𝐫}\right)=\sum _{j=1}^{N_{\mathrm{a}}}{f}_{\mathrm{c}\mathrm{u}\mathrm{t}}\left(r_{ij}\right)g(\mathbf{𝐫}-{\mathbf{𝐫}}_{ij})$

and $g\left(\mathbf{𝐫}\right)$ is an approximation of the delta function. A basis set expansion of ${\rho }_i^{(2)}(r)$ yields the expansion coefficients $c_{n00}^i$, which are used in practice to describe the atomic environment; refer to the theory of machine-learned force fields for details. The tag ML_RCUT1 sets the cutoff radius $R_{\text{cut}}$ at which the cutoff function $f_{\mathrm{c}\mathrm{u}\mathrm{t}}\left(r_{ij}\right)$ decays to zero.

Mind: The cutoff radius determines how many neighbor atoms $N_{\mathrm{a}}$ are considered to describe each central atom's environment. Hence, important features may be missed if the cutoff radius is too small. On the other hand, a large cutoff radius increases the computational cost of the descriptor as the cutoff sphere contains more neighbor atoms. A good compromise is always system-dependent. Therefore, different values should be tested to achieve satisfying accuracy and speed.

The unit of the cut-off radius is $\AA$.

Related tags and articles

ML_LMLFF, ML_RCUT2, ML_W1, ML_SION1, ML_SION2, ML_MRB1, ML_MRB2

Examples that use this tag

---