dill

# generate a random phylogenetic tree, then apply multidimensional
# scaling to the distances and see what the 2D plot looks like

library(ape)

layout(matrix(1:12, 3, 4, byrow=TRUE))

for (i in 1:3){
  # generate a tree
  rr <- rtree(20)

dill

# example using an MRF for ordered categorical predictors

library(mgcv)


# simulate some data...
set.seed(2)
dat <- gamSim(1,n=400,dist="normal",scale=2)

# fit a simple model

dill

library(ggplot2)
library(emoGG)
library(readr)
library(tidyr)
library(dplyr)

# let's visualise the content of season 1 of Bob Ross' The Joy of Painting
dat <- read_csv("https://raw.githubusercontent.com/fivethirtyeight/data/master/bob-ross/elements-by-episode.csv")

# just get what we need and select season 1

dill

dill

# find bad variable/function names in a package
library(stringr)

# retrieve args and functions exported by a package
get_pkgdat <- function(pkgname){

  contents <- objects(paste0("package:", pkgname))

  fns <- Filter(function(x) is.function(eval(parse(text=x))), contents)

dill

  # need to define half-normal point transect detection function pdf
  dhnpt <<- nimbleFunction(
      run = function(x = double(0), b0 = double(0), b1 = double(0),
                     covar = double(0), width = double(0),
                     log = integer(0, default = 0)) {
          returnType(double(0))

        # calculate scale parameter
        sigma <- exp(b0 + b1*covar)

dill

#' USFWS palette generator
#'
#' Palettes taken from https://www.instagram.com/p/CFP2H7JBSxR/
#'
#' @param n Number of colours desired (max 5!). If omitted, uses all colours.
#' @param name one of "sockeye", "coaster", "dolly" or "florida"
#' @param type Either "continuous" or "discrete". Use continuous if you want
#'   to automatically interpolate between colours.
#' @return A vector of colours.
#' @export

dill

# put all yr models in a list
model_list <- list(b_norm, b_term, b_term_sel)
# name them for nice labels
names(model_list) <- c("No selection", "Shrinkage smoother", "Extra penalty")
# all the terms (there's probably a gratia built-in for this)
term_list <- c("s(Depth)", "s(Bottom)", "s(Surface)")

library(gratia)

# pre-storage

dill

# Markov Random Fields with temporal interactions
# based on https://gist.github.com/dill/ecfe7d2f0e542bb274ff

# David L Miller 2020
# Released under MIT license, YMMV


# example from ?mgcv::smooth.construct.mrf.smooth.spec
library(mgcv)
## Load Columbus Ohio crime data (see ?columbus for details and credits)

dill

# extrapolating into the future with B-splines
#  based on code in ?smooth.construct.bs.smooth.spec

library(mgcv)

# annual diameter of women’s skirts at the hem 1866 to 1911
# Hipel and McLeod, 1994
skirts <- scan("http://robjhyndman.com/tsdldata/roberts/skirts.dat",skip=5)
skirtseries <- data.frame(year=1866:1911, diam=skirts)