Douglas Nychka


Manuscripts | Talks | My Calendar | Temp Directory | Staff page


Voice (303) 497-1711, FAX (303) 497-2483 Cell (303) 725-3199    Email nychka "at" ucar "dot" edu

BIO   (text) CV   Everything: (html)  (pdf) (LaTeX) NSF style (but topical): (pdf) (LaTeX)    One pager: (pdf) (LaTeX)



Data Analytics in R

Joint NCAR U Wyoming short course, Boulder, June, 2013

Ten Lectures on Statistics and Climate

CBMS Lecture Series, University of Washington, Seattle, August 2012

Short course on Statistics and Regional Climate

Institute of Mathematical Sciences, National University of Singapore, Feb 28 - March 4 2011

NARCCAP Companion

Useful datasets and source for interpreting the NARCCAP model output..

Daily Surface Ozone Eastern US 1995-1999

An example of a nonstationary spatial process and spatial extremes .


A case study dataset on Colorado climate:

Regional climate model and observed precipitation data for the Colorado Front Range.


R packages

LatticeKrig is a spatial method for large data sets that builds on compactly supported basis functions, Markov random fields and sparse matrix methods.

Downloading LatticeKrig:

fields (home page) is a collection of programs based in R for curve and function fitting with an emphasis on spatial data and flexible covariance functions for Kriging.

Downloading Fields:

Short course CD A directory with the lectures, source code, R binaries and R packages used in the ENAR short course, MAR 2009.

spam is a collection of functions based in R/Fortran for sparse matrix algebra. Written by Reinhard Furrer with the attention to detail that have made the Swiss famous! The fields pacakage uses these functions for spatial analysis of large datasets. Current CRAN version: Version 0.23 SEP-2010.

Major fields functions:
Tps: Thin Plate spline regression
Krig: Spatial process estimate (Kriging)
This function allows you to supply a covariance function as R code, uses sparse matrix methods from the spam package and can handle large data sets.
mKrig (micro Krig ) and fastTps
Fast spatial prediction that can take advantage of compactly supported covariance functions and handle big data sets
cover.design: Finds a space filling design
as.image, image.plot, quilt.plot, crop.image, average.image, designer.colors: Some useful functions for working with image data on 2-d grids and color scales
sreg, qsreg : 1-d smoothing splines and 1-d quantile splines

There are also generic functions that support these methods such as:
plot diagnostic plots of fit
summary statistical summary of fit
surface graphical display of fitted surface
predict, predict.se evaluation fit and prediction error at arbitrary points


nnreg

Neural Networks Package. Estimates a function using a single hidden layer neural network by nonlinear least squares. The fitting algorithm is both robust and accurate. Has supporting functions for diagnostics, GCV and graphing.
Doug Nychka (contact), Stephen Ellner and Barbara Bailey nnreg_1.1.tar.gz (59K)


lenns

Lyapunov Exponents fit by Neural Networks
Fits nonlinear autoregressive maps to multivariate time series data and estimates global and local Lyapunov Exponents. Doug Nychka (contact), Stephen Ellner and Barbara Bailey lenns_1.0.tar.gz (25K)