imputeAlgoPerf

[<code><a rd-options="" href="/link/ASScenario?package=aslib&version=0.1" data-mini-rdoc="aslib::ASScenario">ASScenario</a></code>]
Algorithm selection scenario.

asscenario

[<code>character(1)</code>]
Measure to impute.
Default is first measure in scenario.

measure

[<code>numeric(1)</code>]
See formula.
Default is <code>NULL</code>, which means maximum of performance values if measure should be minimized,
or minimum for maximization case.

base

[<code>numeric(1)</code>]
See formula.
Default is 0.3.

range.scalar

[<code>numeric(1)</code>]
See formula.
Default is 0.

jitter

[<code>logical(1)</code>]
Should values which are exactly 0 be imputed to 1e-6?
This allows to take the logarithm later on, handy for subsequent visualizations.
Note that this really only makes sense for non-negative measures!
Default is FALSE.

impute.zero.vals


The following formula is used for imputation:
<code>base +- range.scalar * range.span + N(0, sd = jitter * range.span)</code>
With <code>range.span = max - min</code>.<p></p><p>Returns an object like <code>algo.runs</code> of <code>asscenario</code>, but drops
the runstatus and all other measures.
</p>

Provides an interface to the algorithm selection benchmark library
at <http://www.aslib.net> and the 'LLAMA' package
(<https://cran.r-project.org/web/packages/llama/index.html>) for building
algorithm selection models.

Lars Kotthoff

aslib

Interface to the Algorithm Selection Benchmark Library

imputeAlgoPerf function

[<code><a rd-options='' href='ASScenario'>ASScenario</a></code>]
Algorithm selection scenario.


The following formula is used for imputation:
<code>base +- range.scalar * range.span + N(0, sd = jitter * range.span)</code>
With <code>range.span = max - min</code>.

Returns an object like <code>algo.runs</code> of <code>asscenario</code>, but drops
the runstatus and all other measures.


imputeAlgoPerf: Imputes algorithm performance for runs which have NA performance values.

Description

Usage

Arguments

Value