evir

The evir package is an R port of Alexander McNeil's S library     
EVIS. The file RCHANGES documents the main changes from the S 
version. As usual, please direct all queries and bug reports 
to the Maintainer given in the DESCRIPTION file.

The functions within the package may be divided into the 
following groups:

Exploratory Data Analysis; these functions provide a useful 
prelude to any further analysis.
  emplot  exindex  hill  meplot  qplot  records

Block Maxima; These functions implement the most classical 
form of EVT analysis.
  gev  gumbel  plot.gev  rlevel.gev

Peaks Over Thresholds; Most users will find these the most 
useful functions.
  gpd  gpd.q  gpd.sfall  plot.gpd  quant  riskmeasures 
  shape  tailplot

Bivariate Peaks Over Thresholds; These functions implement 
a simple form of bivariate EVT.
  gpdbiv  interpret.gpdbiv  plot.gpdbiv

Peaks Over Thresholds with Point Processes; Expert users with 
a deep understanding of the theory might try these functions.
  plot.pot  pot

GEV and GPD Distributions; These functions perform simple 
calculations of densities, probabilities and quantiles, and 
carry out random generation for the generalized extreme value 
and generalized Pareto distributions.
  dgev  dgpd  pgev  pgpd  qgev  qgpd  rgev  rgpd 

Internal; These functions are fully documented, but are 
primarily constructed to be called by other functions.
  findthresh  decluster

The following datasets are also included in the package.
  bmw  danish  nidd.annual  nidd.thresh  siemens  spto87  
  sp.raw

The following four demonstrations, based on a document by 
Alexander McNeil, are included in the package. They can be 
run using the function demo.

  pot.danish       POT Analysis of the Danish Fire Data
  minima.sp        Annual Minima of S&P Daily Return Data
  pot.bmw          POT Analysis of BMW Return Data
  bvpot            Bivariate POT Analysis of BMW and Siemens