collectRectangularMatrix retrieves a distributed rectangular matrix from the slave processes, reconstructing the blocks correctly on the master process. Objects can be copied from environments, lists, and ReferenceClass objects as well as the global environment on the slave processes. WARNING: do not use with a distributed symmetric square matrix; instead use collectTriangularMatrix.
bigGP is a collection of functions for doing distributed calculations in the context of various kinds of Gaussian process models, combined with a ReferenceClass, krigeProblem, for doing kriging calculations based on maximum likelihood estimation.
remoteCrossProdMatSelf multiplies the transpose of a distributed rectangular matrix by itself. remoteCrossProdMatSelf calculates only the diagonal of the crossproduct. The objects can both be contained within environments or ReferenceClass objects as well as the global environment on the slave processes.
distributeVector distributes a vector to the slave processes, breaking into the appropriate pieces, in some cases with padded elements. Objects can be distributed to environments and ReferenceClass objects as well as the global environment on the slave processes.
Copies all objects with a given name from the slave processes to the master process, returning a list with one element per slave process. Objects can be copied from lists, environments, and ReferenceClass objects as well as the global environment on the slave processes.
● Data Source:
CranContrib
● Keywords: get
● Alias: pull
●
0 images
These internal functions carry out the tasks of their respective primary functions, e.g., collectVector, on the slave process. The functions need to be executed on the slave processes.
localGetVectorIndices, localGetTriangularMatrixIndices, and localGetRectangularMatrixIndices are internal auxiliary functions that determine the indices of the elements of a vector or matrix that are stored on a slave process. These are primarily meant for internal use, but can also be used in the process of creating distributed vectors and matrices on the slave processes. The functions need to be executed on the slave processes.
collectVector retrieves a distributed vector from the slave processes, reconstructing in the correct order on the master process. Objects can be copied from environments, lists, and ReferenceClass objects as well as the global environment on the slave processes.
The krigeProblem class provides functionality for kriging using distributed calculations, based on maximum likelihood estimation. The class includes methods for standard kriging calculations and metadata necessary for carrying out the methods in a distributed fashion.