LBFGS_FFM
The LBFGS_FFM object provides a set of functions to interact with the L-BFGS C library shared object through the use of Java's FFM abstractions, allowing the user to call the C implementation of the L-BFGS algorithm directly from Scala code.
Attributes
- Graph
-
- Supertypes
-
class Objecttrait Matchableclass Any
- Self type
-
LBFGS_FFM.type
Members list
Value members
Concrete methods
Copies the data in this LBFGSPrs instance to a destination MemorySegment.
Copies the data in this LBFGSPrs instance to a destination MemorySegment.
It is expected by this method that the destination memory segment was allocated using the memoryLayout MemoryLayout.
This method is only needed for the LBFGS_FFM object implementation. If you are NOT working with LBFGS_FFM, feel free to disregard it.
Value parameters
- destination
-
MemorySegmentthat will receive the data from thisLBFGSPrmsinstance. Must have been allocated using thememoryLayoutMemoryLayout. - prm
-
the LBFGS Parameters
Attributes
L-BFGS main function FFM wrapper for the L-BFGS C library shared object implementation. Calls a C shared object library to perform the L-BFGS optimization that minimizes variables according to the parameters specified by the user.
L-BFGS main function FFM wrapper for the L-BFGS C library shared object implementation. Calls a C shared object library to perform the L-BFGS optimization that minimizes variables according to the parameters specified by the user.
Value parameters
- instanceMemorySegment
-
MemorySegmentwith user data to be provided to theevaluateandprogressmethods. Can be set toMemorySegment.NULLif no user data is required on theevaluateorprogressmethod calls. If not set toMemorySegment.NULLit must be encoded in the sameMemoryLayoutas the one expected by the implementations contained in theevaluateandprogressmethods. - n
-
The number of variables.
- optimizationMethodHandles
-
OptimizationMethodHandlesFFMclass providing the method handles for theevaluateandprogressmethods needed for performing the variable optimization. - params
-
LBFGSPrmsclass representing the parameters chosen to control the L-BFGS optimization. The default parameters used are the defaults of theLBFGSPrmsconstructor. - x
-
VectorDwith the initial values of the variables.
Attributes
- Returns
-
LBFGSResults Results for the L-BFGS optimization. In this implementation, if the objective function is never evaluated due to errors in the arguments from the method call, the
finalFunctionValuereturned will beSome(0). Also, this implementation is currently not capable of determining thelineSearchIncompleteResultsvalue, which will always be set toNoneregardless of how the L-BFGS optimization terminates.