basis_odp

Orthogonal Discriminant Projection (ODP) is a linear dimension reduction
method with class supervision. It maximizes weighted difference between local
and non-local scatter while local information is also preserved by
constructing a neighborhood graph.

Data visualization tours animates linear projection
of multivariate data as its basis (ie. orientation) changes. The 'spinifex'
packages generates paths for manual tours by manipulating the contribution of
a single variable at a time Cook & Buja (1997)
<doi:10.1080/10618600.1997.10474754>. Other types of tours, such as grand
(random walk) and guided (optimizing some objective function) are available
in the 'tourr' package Wickham et al. <doi:10.18637/jss.v040.i02>.
'spinifex' builds on 'tourr' and can render tours with 'gganimate' and
'plotly' graphics, and allows for exporting as an .html widget and as an .gif,
respectively. This work is fully discussed in Spyrison & Cook (2020)
<doi:10.32614/RJ-2020-027>.

Nicholas Spyrison

spinifex

Manual Tours, Manual Control of Dynamic Projections of Numeric
Multivariate Data

Dianne Cook

basis_odp function

<dl><dt>data</dt>
<dd>Numeric matrix or data.frame of the observations, coerced to matrix.</dd>
<dt>class</dt>
<dd>The class for each observation, coerced to a factor.</dd>
<dt>d</dt>
<dd>Number of dimensions in the projection space.
of <code>class</code>.</dd>
<dt>type</dt>
<dd>A vector specifying the neighborhood graph construction.
Expects; <code>c("knn", k)</code>, <code>c("enn", radius)</code>, or <code>c("proportion",ratio)</code>.
Defaults to <code>c("knn", sqrt(nrow(data)))</code>, nearest neighbors equal to the
square root of observations.</dd>
<dt>...</dt>
<dd>Optional, other arguments to pass to <code>Rdimtools::do.odp</code>.</dd></dl>

Arguments

The basis of Orthogonal Discriminant Projection (ODP) — basis_odp

<dl>

<dt>data</dt>
<dd>Numeric matrix or data.frame of the observations, coerced to matrix.</dd>


<dt>class</dt>
<dd>The class for each observation, coerced to a factor.</dd>


<dt>d</dt>
<dd>Number of dimensions in the projection space.
of <code>class</code>.</dd>


<dt>type</dt>
<dd>A vector specifying the neighborhood graph construction.
Expects; <code>c("knn", k)</code>, <code>c("enn", radius)</code>, or <code>c("proportion",ratio)</code>.
Defaults to <code>c("knn", sqrt(nrow(data)))</code>, nearest neighbors equal to the
square root of observations.</dd>


<dt>...</dt>
<dd>Optional, other arguments to pass to <code>Rdimtools::do.odp</code>.</dd>

</dl>

basis_odp: The basis of Orthogonal Discriminant Projection (ODP)

Description

Usage

Arguments

References

See Also

Examples