| Type: | Package |
| Title: | Partitioned Distance Function |
| Version: | 1.2.1 |
| Date: | 2013-01-31 |
| Author: | Jeffrey Wong |
| Maintainer: | Jeffrey Wong <jeff.ct.wong@gmail.com> |
| URL: | https://github.com/jeffwong/pdist |
| Description: | Computes the euclidean distance between rows of a matrix X and rows of another matrix Y. Previously, this could be done by binding the two matrices together and calling 'dist', but this creates unnecessary computation by computing the distances between a row of X and another row of X, and likewise for Y. pdist strictly computes distances across the two matrices, not within the same matrix, making computations significantly faster for certain use cases. |
| License: | GPL-2 | GPL-3 [expanded from: GPL] |
| LazyLoad: | yes |
| Depends: | methods |
| Collate: | 'pdist.R' 'setup.R' |
| Packaged: | 2022-05-02 10:08:08 UTC; hornik |
| Repository: | CRAN |
| Date/Publication: | 2022-05-02 13:01:57 UTC |
| NeedsCompilation: | yes |
Distances between Observations for a Partitioned Matrix
Description
Computes the euclidean distance between rows of a matrix X and rows of another matrix Y. Previously, this could be done by binding the two matrices together and calling 'dist', but this creates unnecessary computation by computing the distances between a row of X and another row of X, and likewise for Y. pdist strictly computes distances across the two matrices, not within the same matrix, making computations significantly faster for certain use cases.
Details
| Package: | pdist |
| Type: | Package |
| Version: | 1.2 |
| Date: | 2013-01-31 |
| License: | GPL |
| LazyLoad: | yes |
Author(s)
Jeffrey Wong <jeff.ct.wong@gmail.com>
extract parts of pdist
Description
extract parts of pdist
Partitioned Distances
Description
Compute distance matrix between two matrices of observations, or two subsets of one matrix
Usage
pdist(X, Y = NULL, indices.A = NULL, indices.B = NULL)
Arguments
X |
a matrix of n observations where columns represent features of the observations |
Y |
optional. A second matrix of p observations like X. Y must have the same number of columns as X |
indices.A |
optional. A vector of integers representing row indices from X. This should only be used when Y is not provided. |
indices.B |
optional. A vector of integers representing row indices from X. This should only be used when Y is not provided. |
Details
pdist computes a n by p distance matrix using two seperate matrices. pdist allows the user to factor out observations into seperate matrices to improve computations. The function dist computes the distances between all possible pair wise elements, pdist only computes the distance between observations in X with observations in Y; distances between observations in X and other observations in X are not computed, and likewise for Y.
If a second matrix Y is not provided, indices.A and indices.B can be provided together to specify subsets of X to be computed. A new matrix X is created by taking X[indices.A,] and Y is created using X[indices.B,].
The return value of pdist is a distance vector, much like the default return value for dist. However, it can be accessed like a full distance matrix. If mypdist = pdist(X,Y), mypdist[i,j] is the distance between X[i,] and Y[j,]. Similarly, mypdist[i,] is a vector of distances between X[i,] and all observations in Y.
Examples
x = matrix(rnorm(100),10,10)
x.pdist = pdist(x, indices.A = 1:3, indices.B = 8:10)
message("Find the distance between observation 1 and 10 of x")
x.pdist[1,3]
message("Converting a pdist object into a traditional distance matrix")
as.matrix(x.pdist)