impulse.fit.double: Fit Impulse Model (Double sigmoid)

In farrellja/URD: URD

Description

Attempts to fit a double sigmoid ("impulse" model) to given expression data using expectation maximization.

Usage

impulse.fit.double(x, y, k = 20, maxiter = 200, interpolate = NULL)

Arguments

x	(Numeric) Expression data
y	(Numeric) Expression data
k	(Numeric) Number of starting conditions to try
maxiter	(Numeric) Maximum number of iterations to try fitting
interpolate	(Numeric or NULL) If low number of data points, can interpolate them linearly to this number of points for choosing potential starting conditions. Default (NULL) does not do any interpolation. Interpolation is NOT used during actual function fitting.

Value

(Named vector, length 8) Returns the coefficients and sum of squared residuals of the best fitting double sigmoid out of k conditions. If all attempts to fit fail to converge, returns vector of same length, with all values NA.

farrellja/URD documentation built on June 17, 2020, 4:48 a.m.