Difference between revisions of "Optimize"

From TurboGAP
Jump to navigation Jump to search
(Created page with "<code>optimize</code> selects the method for propagating the atomic positions. Currently, the accepted options are: # <code>"vv"</code> for Velocity Verlet molecular dy...")
 
 
(3 intermediate revisions by one other user not shown)
Line 1: Line 1:
 
<code>[[optimize]]</code> selects the method for propagating the atomic positions. Currently, the accepted options are:
 
<code>[[optimize]]</code> selects the method for propagating the atomic positions. Currently, the accepted options are:
  
# <code>"vv"</code> for Velocity Verlet [[molecular dynamics]] (this is the default if you don't choose anything)
+
# <code>"vv"</code> for [[Velocity Verlet]] [[molecular dynamics]] (this is the default if you don't choose anything)
 
# <code>"gd"</code> for [[gradient descent]] [[energy minimization]]
 
# <code>"gd"</code> for [[gradient descent]] [[energy minimization]]
  
Line 16: Line 16:
 
| Optional
 
| Optional
 
| String
 
| String
| <code>"vv"</code> or <code>[[gradient descent|gd]]</code>
+
| <code>"[[velocity verlet|vv]]"</code> or <code>"[[gradient descent|gd]]"</code>
 
| <code>"gd"</code>
 
| <code>"gd"</code>
 
| <code>[[max_opt_step]]</code>
 
| <code>[[max_opt_step]]</code>
Line 23: Line 23:
 
=== Example ===
 
=== Example ===
  
  [[optimize]] = "gd"
+
  [[optimize]] = "gd" ! Do gradient descent minimization in this case
 +
[[optimize_groupID]] = all ! Must mention if there is a particular group of atoms, otherwise 'all' is default and need not be mentioned.
 +
[[e_tol]] = 1.e-6 ! Energy threshold for convergence
 +
[[md_nsteps]] = 1000 ! Maximum number of iterations (it will break earlier if convergence to e_tol is reached)
 
  [[max_opt_step]] = 0.1 ! Angstrom
 
  [[max_opt_step]] = 0.1 ! Angstrom

Latest revision as of 13:17, 6 November 2023

optimize selects the method for propagating the atomic positions. Currently, the accepted options are:

  1. "vv" for Velocity Verlet molecular dynamics (this is the default if you don't choose anything)
  2. "gd" for gradient descent energy minimization

Summary

Summary for optimize keyword
Required/optional Type Accepted values Default See also
Optional String "vv" or "gd" "gd" max_opt_step

Example

optimize = "gd" ! Do gradient descent minimization in this case
optimize_groupID = all ! Must mention if there is a particular group of atoms, otherwise 'all' is default and need not be mentioned.
e_tol = 1.e-6 ! Energy threshold for convergence
md_nsteps = 1000 ! Maximum number of iterations (it will break earlier if convergence to e_tol is reached)
max_opt_step = 0.1 ! Angstrom