Description Usage Arguments Value Methods Author(s) See Also Examples

Generation of Multidimensional Scaling objects for the dissimilarities
between elements given as an input in a `distGPS`

object. Metric
and non-metric algorithms are available, as well as an optimization
algorithm for improving r-square correlation between observed and
approximated distances. The MDS calculation for a given distance
matrix can be splitted into smaller individual tasks and run in
parallel, greatly improving CPU time and system memory usage. The S4
accessor functions `getR2, getStress, getPoints`

retrieve R-square correlation, stress and points stored within a
`mds`

object respectively. The function `is.adj`

is useful
to know if a certain chroGPS MDS map has been adjusted by Procrustes
or not (see help for `procrustesAdj`

for details.)

1 2 3 4 |

`d` |
Object of class |

`m` |
(Optional). Object of class |

`k` |
Dimensionality of the reconstructed space, typically set to 2 or 3. |

`type` |
Set to |

`add` |
Logical indicating if an additive constant c* should be
computed, and added to the non-diagonal dissimilarities such
that all n-1 eigenvalues are non-negative in |

`cor.method` |
A character string indicating which correlation coefficient (or covariance) is to be computed. One of "pearson" (default), "kendall", or "spearman", can be abbreviated. |

`splitMDS` |
Set to |

`split` |
Proportion of elements to include in each (but last) distance submatrix. |

`overlap` |
Proportion of elements to be used as common anchor points between two adjacent distance submatrixes. These points will be used as spatial references to stitch each two adjacent MDS objects by Procrustes. |

`stepSize` |
Size for the quadratic search step to be used for R-square optimization if |

`reshuffle` |
Set to TRUE to perform random resampling of the input distance matrix before splitting it for parallel computation. This is often necessary to sufficiently capture the inherent variability of the data in each distance submatrix so that the stitching process can work properly, as the original data may present an arbitrary sorting of some kind. If a previous resampling of the data has been performed, this is not necessary. |

`set.seed` |
Random seed to perform the resampling. |

`mc.cores` |
Number of cores to be passed to the |

`...` |
Additional parameters passed to |

The function returns a `mds`

object. See help ("mds-Class") for details.

- mds
`signature(d = "distGPS", m = "missing")`

: Creates a`mds`

object with points in a k-dimensional space approximating the pairwise distances in`d`

.- mds
`signature(d = "distGPS", m = "mds")`

: For the observed dissimilarities in`d`

and a valid spatial representation of them in`m`

, the function returns a`mds`

object with an optimized representation of`d`

in terms of R-square. The MDS stress measure is also returned. See help for`boostMDS`

for details.- plot
`signature(m = "mds")`

: S4 plot method for`mds`

objects.

Oscar Reina

See functions cmdscale, isoMDS from package `MASS`

.

1 2 3 4 5 6 7 |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.