Skip to content

Instantly share code, notes, and snippets.

@dennisprangle
dennisprangle / jags_install.md
Last active March 20, 2023 14:10
JAGS installation notes
@dennisprangle
dennisprangle / beamerthemeNCLDennis.sty
Created May 12, 2019 15:46
A Newcastle University beamer theme
% adapted Philippe Dreuw and Thomas Deselaers
% further adapted from edits by James Malcolm
% finally, Shawn Lankton made some edits (www.shawnlankton.com)
% and I added some quick changes - Dennis Prangle
\ProvidesPackage{beamerthemeNCLDennis}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\mode<presentation>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
@dennisprangle
dennisprangle / SGD.R
Last active November 4, 2019 11:00
R implementation of stochastic gradient descent algorithms
# Reference class objects for gradient descent algorithms.
# The "step" method take an estimated gradient as argument
# and returns an increment to apply to the current parameter values.
# See end of file for an example.
##Algorithm which does no adaptation
noAdapt = setRefClass("noAdapt",
fields = list(),
methods = list(
step = function(gradientEst) {