Objective functions giving the likelihoods computed by
likeLTD.
These functions do not exist per se. They are created by
create.likelihood
and similar function.
The general workflow of likeLTD will generally result in the creation and maximization of one or more objective function. These functions will take the arguments listed above. They can also be accept any number of other arguments. This makes it easy both to extend likeLTD and to use objective functions with penalty functions.
Some of the arguments are verified for size and/or value.
The general form of these objective functions is:
function(locusAdjustment, power, dropout, degradation=NULL,
rcont=NULL, dropin=NULL, ...)
locusAdjustmentA scalar (single locus objective) or a list of values giving the locus adjustment to dropout for each locus. Must be positive.
powerTvedebrink exponent. Must be a scalar and negative.
dropoutDropout rate for each replicate. Must be between 0 and 1.
degradationDegradation parameters to determine likeliness of dropout. Should be of length k + u.
rcontRelative contribution. Must be of length k + u or k
+ u - 1, where k is the number of contributors with
known profiles, and u is the number of unknown
contributors. Since the input is for relative
contributions, there are two possible sizes. In the first
case, the contribution of a reference individual is kept to 1.
In general, the reference individual will be the queried
individual (if subjecto dropout and in prosecution hypothesis)
or the first individual subject to dropout.
In the second case, the contributions are used for all
known and unknown contributors as given. It is for the user
to know what she is doing. rcont
should always be
positive.
dropinA scalar giving the dropin rate. Ignored in models without dropin. Otherwise it should be positive (unless given in exponential form).
... Any other parameter. Mostly ignored by objective functions, but it makes it easier to pass on parameters to the penalty function
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.