The R package
glamlasso
fits 3D generalized linear array models with an l1-penalty. It exploits the tensor
product structure of the design matrix for array models via a proximal gradient algorithm,
which makes it easy and fast to fit 3D array models.

The implemented algorithm is described in
Lund et al. (2016), which
also contains examples of applications and benchmark studies.

The R package ppstat (point process statistics)
fits multivariate point processes. The package processdata
implements data structures used for ppstat and various plotting and data management features.

The methods for multivariate Hawkes processes were used in Carstensen et al. (2010).
Some theory is found in Hansen (2013), Hansen (2014)
and Hansen, Reynaud-Bouret and Rivoirard (2014).

See the ppstat page for more information.

Shiny applications for experimenting with classical probability results (LLN, CLT etc) are available from
the GitHub repository shinyProbability.

The R package msgl (multinomial sparse group lasso)
fits multinomial models for e.g. multiclass classification with a sparse group lasso penalty. It is based on the generic sparse group
lasso algorithm implemented in the R package sglOptim.

The generic algorithm as well as the multinomial application is described in Vincent and Hansen (2014).
It is applied for a multiclass classification problem in Vincent et al. (2014). See also the
R code used in the latter paper for more information.

The R package smde (sparse multivariate differential equations)
fits multivariate and high-dimensional ODE and SDE models using l1-penalized least squares estimation. It contains
an implementation of a generic algorithm for l1-penalized nonlinear least squares problems.

The package was used in Hansen and Sokol (2014) in a simulation study to
illustrate the general results on degrees of freedom when fitting a linear multivariate ODE using nonlinear least
squares with an l1-contraint.

The R package expoRkit implements an R interface to the
Fortran package Expokit developed by Roger B. Sidje.

Expokit is an efficient Fortran implementation for computing the
matrix exponential, or rather, its action on a vector, for large
sparse matrices. This can also be understood as computing the
solution of a system of linear ordinary first order differential
equations.

The program

*StemSearch*is a command-line program implemented in C++ that can scan a genome sequence for putative stem-loop structures. The output contains a ranked list of high-scoring putative stem-loops with a normalized nat-score and an E-value to guide the assessment of statistical significance.
StemSearch is a dedicated datamining program, and effort is made to
discriminate structures from the bulk
genome and to give a proper statistical
treatment of the results. The program was used in Hansen (2009),
which treats the theory behind the statistics.

To use the program, download
the gnuzipped tar-file (current version: 0.9, released July 1, 2008),
gunzip into an appropriate directory,
untar, read and follow the instructions in the
file INSTALL.