Gpr-opthyprms: Difference between revisions

From PMFLib Wiki
Jump to navigation Jump to search
No edit summary
 
No edit summary
 
Line 2: Line 2:
Navigation: [[Documentation]] / [[Utilities]] / [[gpr-opthyprms]]
Navigation: [[Documentation]] / [[Utilities]] / [[gpr-opthyprms]]
----
----
<div style="text-align: center">gpr-opthyprms: version 6.380.1d87074 (2021-07-26 16:36)</div>'''Name'''
<div style="text-align: center">gpr-opthyprms: version 6.1090.a7cf688 (2026-06-18 21:15)</div>'''Name'''


<span style="margin-left: 15px;">gpr-opthyprms</span>
<span style="margin-left: 15px;">gpr-opthyprms</span>
Line 8: Line 8:
'''Synopsis'''
'''Synopsis'''


<span style="color: maroon; margin-left: 15px;">gpr-opthyprms</span> <span style="color: blue;">[OPTIONS]</span> [--] <span style="color: purple;">accuname1 realm1 [accuname2 realm2 [..]] hyprmsname</span>
<span style="color: maroon; margin-left: 15px;">gpr-opthyprms</span> <span style="color: blue;">[OPTIONS]</span> [--] <span style="color: purple;">ACCU</span> <span style="color: purple;">REALM</span> <span style="color: purple;">HYPRMS</span>


'''Description'''
'''Description'''
Line 16: Line 16:
'''Arguments'''
'''Arguments'''


<div style="margin-left: 2em;"><span style="color: darkcyan;"><span style="font-style: bold;">accuname1</span></span>                 Name of file containing the PMF accumulator.<br /><span style="color: darkcyan;"><span style="font-style: bold;">realm1</span></span>                     Realm of interest.<br /><span style="color: darkcyan;"><span style="font-style: bold;">hyprmsname</span></span>                 Name of file where the optimized GPR hyperparameters are saved. If the name is '-' then the output will be written to the standard output.</div>
{| style="margin-left: 2em; width: 95%;"
|- style="vertical-align: top; background-color: #e6e6e6;"
| <span style="color: purple;">ACCU</span><br/><div style="margin-left: 30px;">Name of the file containing the input PMF accumulator.</div>
|- style="vertical-align: top;"
| <span style="color: purple;">REALM</span><br/><div style="margin-left: 30px;">Realm for GPR hyperparameter optimization. The list of supported realms can be obtained by --listrealms.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
| <span style="color: purple;">HYPRMS</span><br/><div style="margin-left: 30px;">Name of the file containing the optimized hyperparameters.</div>
|}


'''Options'''
'''Options'''
Line 22: Line 29:
{| style="margin-left: 2em; width: 95%;"
{| style="margin-left: 2em; width: 95%;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">-t, --target NAME</span><br/><div style="margin-left: 30px;">Specify optimized targed, which can be either logml (log of marginal likelihood) or logpl (log of pseudo-likelihood from leave-one-out cross-validation (LOO-CV)).</div><span style="color:purple; margin-left: 30px;">(Default: logml)</span>
|<span style="color: blue;">--listrealms</span><br/><div style="margin-left: 30px;">List supported realms.</div>
|- style="vertical-align: top;"
|<span style="color: blue;">-t, --target NAME</span><br/><div style="margin-left: 30px;">Specify the optimized target, which can be either logml (log of marginal likelihood) or logpl (log of pseudo-likelihood from leave-one-out cross-validation, LOO-CV).</div><span style="color:purple; margin-left: 30px;">(Default: logml)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">-l, --limit LIMIT</span><br/><div style="margin-left: 30px;">Only bins containing more samples than NUMBER are considered properly sampled.</div><span style="color:purple; margin-left: 30px;">(Default: 1000)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">-a, --lmethod NAME</span><br/><div style="margin-left: 30px;">Linear algebra method for matrix inversion. Supported algorithms are: default, svd (SVD - singular value decomposition, divide and conquer driver), svd2 (SVD - singular value decomposition, simple driver), lu (LU factorization), and ll (LL - Cholesky factorization).</div><span style="color:purple; margin-left: 30px;">(Default: default)</span>
|<span style="color: blue;">-a, --lmethod NAME</span><br/><div style="margin-left: 30px;">Linear algebra method for matrix inversion. Supported algorithms: default, svd (SVD – singular value decomposition, divide-and-conquer driver), svd2 (SVD – singular value decomposition, simple driver), lu (LU factorization), and ll (LL – Cholesky factorization).</div><span style="color:purple; margin-left: 30px;">(Default: default)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">-r, --rcond NUMBER</span><br/><div style="margin-left: 30px;">Rank condition for SVD. Used value must be carefully tested. Calculation at computer precision is requested with -1 (not recommended).</div><span style="color:purple; margin-left: 30px;">(Default: 1e-06)</span>
|<span style="color: blue;">-r, --rcond NUMBER</span><br/><div style="margin-left: 30px;">Rank condition for SVD. The chosen value must be carefully tested. Calculation at computer precision is requested with -1 (not recommended).</div><span style="color:purple; margin-left: 30px;">(Default: 1e-06)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">-s, --sigmaf2 NUMBER</span><br/><div style="margin-left: 30px;">Variance of the reconstructed free energy surface (signal variance).</div><span style="color:purple; margin-left: 30px;">(Default: 15)</span>
|<span style="color: blue;">-s, --sigmaf2 SPEC</span><br/><div style="margin-left: 30px;">Variance of the reconstructed free-energy surface (signal variance) in the form SigmaF2(1)[xSigmaF2(2)[x...]]. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: 15.0)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">-c, --ncorr NUMBER</span><br/><div style="margin-left: 30px;">Number of statistically correlated samples.</div><span style="color:purple; margin-left: 30px;">(Default: 1)</span>
|<span style="color: blue;">--minsigmaf2 NUMBER</span><br/><div style="margin-left: 30px;">Minimum value of SigmaF2.</div><span style="color:purple; margin-left: 30px;">(Default: 0.1)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">-w, --wfac SPEC</span><br/><div style="margin-left: 30px;">Factors influencing widths of RBFs or square exponential kernels. The width is distance between the adjacent square exponential functions multiplied by this factors in the form WFac1[xWFac2x...]. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: 3.0)</span>
|<span style="color: blue;">-o, --covar SPEC</span><br/><div style="margin-left: 30px;">Covariances between GPR tasks in the form CoVar1[xCoVar2[x...]]. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: 0.0)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--enablesigmaf2</span><br/><div style="margin-left: 30px;">Enable optimization of SigmaF2 hyperparameter.</div>
|<span style="color: blue;">--mincovar NUMBER</span><br/><div style="margin-left: 30px;">Minimum value of CoVar.</div><span style="color:purple; margin-left: 30px;">(Default: -1000)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--enablencorr</span><br/><div style="margin-left: 30px;">Enable optimization of NCorr hyperparameter.</div>
|<span style="color: blue;">-w, --wfac SPEC</span><br/><div style="margin-left: 30px;">Characteristic scale of collective variables in the form WFac1[xWFac2[x...]]. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: 3.0)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--enablewfac SPEC</span><br/><div style="margin-left: 30px;">Enable optimization of Wfac hyperparamters. Flags are specified in the form WFac1Enabled[xWFac2Enabledx...] with F and T for disabled and enabled, respectively. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: F)</span>
|<span style="color: blue;">--minwfac NUMBER</span><br/><div style="margin-left: 30px;">Minimum value of WFac.</div><span style="color:purple; margin-left: 30px;">(Default: 0.1)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--numeric</span><br/><div style="margin-left: 30px;">Use numeric gradients.</div>
|<span style="color: blue;">-n, --sigman2 SPEC</span><br/><div style="margin-left: 30px;">Noise variances (σ²) for each CV in the form SigmaN2(1)[xSigmaN2(2)[x...]]. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: 1e-5)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--minsigman2 NUMBER</span><br/><div style="margin-left: 30px;">Minimum value of SigmaN2.</div><span style="color:purple; margin-left: 30px;">(Default: 0)</span>
|- style="vertical-align: top;"
|<span style="color: blue;">--enablesigmaf2 SPEC</span><br/><div style="margin-left: 30px;">Enable optimization of SigmaF2 hyperparameters. Flags are specified in the form SigmaF2(1)Enabled[xSigmaF2(2)Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: F)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--enablecovar SPEC</span><br/><div style="margin-left: 30px;">Enable optimization of CoVar hyperparameters. Flags are specified in the form CoVar1Enabled[xCoVar2Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: F)</span>
|- style="vertical-align: top;"
|<span style="color: blue;">--enablewfac SPEC</span><br/><div style="margin-left: 30px;">Enable optimization of WFac hyperparameters. Flags are specified in the form WFac1Enabled[xWFac2Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: F)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--enablesigman2 SPEC</span><br/><div style="margin-left: 30px;">Enable optimization of SigmaN2 hyperparameters. Flags are specified in the form SigmaN2(1)Enabled[xSigmaN2(2)Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.</div><span style="color:purple; margin-left: 30px;">(Default: F)</span>
|- style="vertical-align: top;"
|<span style="color: blue;">--numeric</span><br/><div style="margin-left: 30px;">Use numerical gradients.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--noptsteps NUMBER</span><br/><div style="margin-left: 30px;">Maximum number of optimization steps.</div><span style="color:purple; margin-left: 30px;">(Default: 100)</span>
|<span style="color: blue;">--noptsteps NUMBER</span><br/><div style="margin-left: 30px;">Maximum number of optimization steps.</div><span style="color:purple; margin-left: 30px;">(Default: 100)</span>
Line 46: Line 69:
|<span style="color: blue;">--nresets NUMBER</span><br/><div style="margin-left: 30px;">Maximum number of resets due to insufficient optimization progress.</div><span style="color:purple; margin-left: 30px;">(Default: 3)</span>
|<span style="color: blue;">--nresets NUMBER</span><br/><div style="margin-left: 30px;">Maximum number of resets due to insufficient optimization progress.</div><span style="color:purple; margin-left: 30px;">(Default: 3)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--termeps NUMBER</span><br/><div style="margin-left: 30px;">Termination criteria for L-BFGS optimizer (see L-BFGS code).</div><span style="color:purple; margin-left: 30px;">(Default: 1e-05)</span>
|<span style="color: blue;">--termeps NUMBER</span><br/><div style="margin-left: 30px;">Termination criterion for the L-BFGS optimizer (see L-BFGS code).</div><span style="color:purple; margin-left: 30px;">(Default: 1e-05)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--termval NUMBER</span><br/><div style="margin-left: 30px;">Termination criteria for L-BFGS optimizer. Minimum change of optimized property.</div><span style="color:purple; margin-left: 30px;">(Default: 1e-06)</span>
|<span style="color: blue;">--termval NUMBER</span><br/><div style="margin-left: 30px;">Termination criterion for the L-BFGS optimizer. Minimum change of the optimized property.</div><span style="color:purple; margin-left: 30px;">(Default: 1e-06)</span>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--nlbfgscorr NUMBER</span><br/><div style="margin-left: 30px;">Number of corrections used in an L-BFGS update.</div><span style="color:purple; margin-left: 30px;">(Default: 10)</span>
|<span style="color: blue;">--nlbfgscorr NUMBER</span><br/><div style="margin-left: 30px;">Number of corrections used in an L-BFGS update.</div><span style="color:purple; margin-left: 30px;">(Default: 10)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--test</span><br/><div style="margin-left: 30px;">Compare analytical and numerical gradient for input hyperparameters.</div>
|<span style="color: blue;">--test</span><br/><div style="margin-left: 30px;">Compare analytical and numerical gradients for input hyperparameters.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--cd5</span><br/><div style="margin-left: 30px;">Use 5-point stencil for numerical differentiation.</div>
|<span style="color: blue;">--cd5</span><br/><div style="margin-left: 30px;">Use the 5-point stencil for numerical differentiation.</div>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--loadhyprms NAME</span><br/><div style="margin-left: 30px;">Name of file with GPR hyperparameters.</div>
|<span style="color: blue;">--loadhyprms NAME</span><br/><div style="margin-left: 30px;">Name of the file containing the GPR hyperparameters.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--stat</span><br/><div style="margin-left: 30px;">Calculate detailed GPR status employing found hyperparameters.</div>
|<span style="color: blue;">--stat</span><br/><div style="margin-left: 30px;">Calculate detailed GPR statistics using the optimized hyperparameters.</div>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--sptype</span><br/><div style="margin-left: 30px;">Determine type of stationary point.</div>
|<span style="color: blue;">--sptype</span><br/><div style="margin-left: 30px;">Determine the type of stationary point.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--kernel NAME</span><br/><div style="margin-left: 30px;">GPR: Kernel type. Supported types: ardse (ARD squared exponential), ardmc52 (ARD Matern class 5/2), default(=ardse)</div><span style="color:purple; margin-left: 30px;">(Default: default)</span>
|<span style="color: blue;">--kernel NAME</span><br/><div style="margin-left: 30px;">GPR kernel type. Supported types: ardse (ARD squared exponential), ardmc52 (ARD Matérn class 5/2), default (=ardse).</div><span style="color:purple; margin-left: 30px;">(Default: default)</span>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--useinv</span><br/><div style="margin-left: 30px;">Use matrix inversion pathway (for testing only).</div>
|<span style="color: blue;">--fdkernel</span><br/><div style="margin-left: 30px;">Use the first derivatives of the GPR kernel (only for SmootherGPR).</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--useinv</span><br/><div style="margin-left: 30px;">Use the matrix inversion pathway (for testing only).</div>
|- style="vertical-align: top;"
|<span style="color: blue;">--calclogpl</span><br/><div style="margin-left: 30px;">Calculate logPL for --stat if --target is not logpl.</div>
|<span style="color: blue;">--calclogpl</span><br/><div style="margin-left: 30px;">Calculate logPL for --stat if --target is not logpl.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">--globalmin SPEC</span><br/><div style="margin-left: 30px;">Position of the global minimum provided as a single string in the form CV1xCV2x...xCVn (relevant for error determination). If not set, the position is determined automatically.</div>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--globalmin SPEC</span><br/><div style="margin-left: 30px;">position of global minimum provided as a single string in the form CV1xCV2x...xCVn (relevant for error determination), if not set the position is determined automatically.</div>
|<span style="color: blue;">-v, --verbose</span><br/><div style="margin-left: 30px;">Increase output verbosity.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">-v, --verbose</span><br/><div style="margin-left: 30px;">Increase output verbosity.</div>
|<span style="color: blue;">--version</span><br/><div style="margin-left: 30px;">Output version information and exit.</div>
|- style="vertical-align: top;"
|- style="vertical-align: top;"
|<span style="color: blue;">--version</span><br/><div style="margin-left: 30px;">Output version information and exit.</div>
|- style="vertical-align: top; background-color: #e6e6e6;"
|<span style="color: blue;">-h, --help</span><br/><div style="margin-left: 30px;">Display this help and exit.</div>
|<span style="color: blue;">-h, --help</span><br/><div style="margin-left: 30px;">Display this help and exit.</div>
|}
|}

Latest revision as of 11:50, 19 June 2026

Navigation: Documentation / Utilities / gpr-opthyprms


gpr-opthyprms: version 6.1090.a7cf688 (2026-06-18 21:15)

Name

gpr-opthyprms

Synopsis

gpr-opthyprms [OPTIONS] [--] ACCU REALM HYPRMS

Description

The program finds optimal GPR hyperparameters, which maximize logarithm of marginal likelihood. The optimization is performed by the L-BFGS optimizer employing either analytical or numerical gradients of logML w.r.t. hyperparameters.

Arguments

ACCU
Name of the file containing the input PMF accumulator.
REALM
Realm for GPR hyperparameter optimization. The list of supported realms can be obtained by --listrealms.
HYPRMS
Name of the file containing the optimized hyperparameters.

Options

--listrealms
List supported realms.
-t, --target NAME
Specify the optimized target, which can be either logml (log of marginal likelihood) or logpl (log of pseudo-likelihood from leave-one-out cross-validation, LOO-CV).
(Default: logml)
-l, --limit LIMIT
Only bins containing more samples than NUMBER are considered properly sampled.
(Default: 1000)
-a, --lmethod NAME
Linear algebra method for matrix inversion. Supported algorithms: default, svd (SVD – singular value decomposition, divide-and-conquer driver), svd2 (SVD – singular value decomposition, simple driver), lu (LU factorization), and ll (LL – Cholesky factorization).
(Default: default)
-r, --rcond NUMBER
Rank condition for SVD. The chosen value must be carefully tested. Calculation at computer precision is requested with -1 (not recommended).
(Default: 1e-06)
-s, --sigmaf2 SPEC
Variance of the reconstructed free-energy surface (signal variance) in the form SigmaF2(1)[xSigmaF2(2)[x...]]. The last value pads the rest.
(Default: 15.0)
--minsigmaf2 NUMBER
Minimum value of SigmaF2.
(Default: 0.1)
-o, --covar SPEC
Covariances between GPR tasks in the form CoVar1[xCoVar2[x...]]. The last value pads the rest.
(Default: 0.0)
--mincovar NUMBER
Minimum value of CoVar.
(Default: -1000)
-w, --wfac SPEC
Characteristic scale of collective variables in the form WFac1[xWFac2[x...]]. The last value pads the rest.
(Default: 3.0)
--minwfac NUMBER
Minimum value of WFac.
(Default: 0.1)
-n, --sigman2 SPEC
Noise variances (σ²) for each CV in the form SigmaN2(1)[xSigmaN2(2)[x...]]. The last value pads the rest.
(Default: 1e-5)
--minsigman2 NUMBER
Minimum value of SigmaN2.
(Default: 0)
--enablesigmaf2 SPEC
Enable optimization of SigmaF2 hyperparameters. Flags are specified in the form SigmaF2(1)Enabled[xSigmaF2(2)Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.
(Default: F)
--enablecovar SPEC
Enable optimization of CoVar hyperparameters. Flags are specified in the form CoVar1Enabled[xCoVar2Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.
(Default: F)
--enablewfac SPEC
Enable optimization of WFac hyperparameters. Flags are specified in the form WFac1Enabled[xWFac2Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.
(Default: F)
--enablesigman2 SPEC
Enable optimization of SigmaN2 hyperparameters. Flags are specified in the form SigmaN2(1)Enabled[xSigmaN2(2)Enabled[x...]] with F and T for disabled and enabled, respectively. The last value pads the rest.
(Default: F)
--numeric
Use numerical gradients.
--noptsteps NUMBER
Maximum number of optimization steps.
(Default: 100)
--nresets NUMBER
Maximum number of resets due to insufficient optimization progress.
(Default: 3)
--termeps NUMBER
Termination criterion for the L-BFGS optimizer (see L-BFGS code).
(Default: 1e-05)
--termval NUMBER
Termination criterion for the L-BFGS optimizer. Minimum change of the optimized property.
(Default: 1e-06)
--nlbfgscorr NUMBER
Number of corrections used in an L-BFGS update.
(Default: 10)
--test
Compare analytical and numerical gradients for input hyperparameters.
--cd5
Use the 5-point stencil for numerical differentiation.
--loadhyprms NAME
Name of the file containing the GPR hyperparameters.
--stat
Calculate detailed GPR statistics using the optimized hyperparameters.
--sptype
Determine the type of stationary point.
--kernel NAME
GPR kernel type. Supported types: ardse (ARD squared exponential), ardmc52 (ARD Matérn class 5/2), default (=ardse).
(Default: default)
--fdkernel
Use the first derivatives of the GPR kernel (only for SmootherGPR).
--useinv
Use the matrix inversion pathway (for testing only).
--calclogpl
Calculate logPL for --stat if --target is not logpl.
--globalmin SPEC
Position of the global minimum provided as a single string in the form CV1xCV2x...xCVn (relevant for error determination). If not set, the position is determined automatically.
-v, --verbose
Increase output verbosity.
--version
Output version information and exit.
-h, --help
Display this help and exit.