Gauss Procedures
This page contains Gauss statistical, utility, and graphic
procedures. Some are actively used, but others have not been updated
or checked in some time. I'd appreciate if you'd let me know if you find any bugs,
have found more efficient or general approaches, or have others to
add. Larger packages I've written are available at my statistical software page; code written by
others can be found at Gauss
resources. You may prefer a zip file
containing all the procedures below (last update: 8/20/2005), or see
below to receive automatic notification of updates of this page.
Statistics
- Amelia, A package for imputing
missing values
- cdfNorm, pdfNorm
unstandardized CDF and PDF of the normal
- cMeanc, cMeanWc, meanSE, conditional means and other statistics
- dens, Kernel density estimation
- EI, a package for ecological inference
- fisherZ, fisherZi, Fisher's Z transformation and its inverse
- GEC, estimate the generalized event count
model
- hyperg, hypergeometric pmf
- lcdfnormI, lncdfn2 for an unstandarized
normal
- logit, logitI, Mlogit, MlogitI,
univariate and multivariate logistic transformation and inverse
- lpdfNorm, lnpdfMN, lnpdfmt, log of
the univariate and multivariate normal, and the Student's t pdfs
- MaxLik, various MLE models and
sample datasets
- meanWc, weighted means (handles missings well)
- PDFs, David Baird's PDFs, CDFs, random numbers
- reg, regR, regV, regVr, regression, with
linear restrictions, and both with Variable name entry
- regRand, regression with approximate randomization
- ReLogit, ReLogitQ, Rare Events Logistic Regression, logit
with corrections for rare events and selection on Y
- rndCHI, random chi-square draws
- rndMN, rndMT, random
(unstandardized) multivariate normal and Student's t draws
- rndSmpl, rndSmpl2, random samples with or without
replacement
- sechol, shows how to take an approximate
Cholesky decomposition for almost any matrix. Useful for optimization
routines when the Hessian is not invertable. See Jeff Gill and Gary
King. "What to do When Your
Hessian is Not Invertible: Alternatives to Model Respecification in
Nonlinear Estimation," Sociological Methods and
Research, Vol. 32, No. 1 (2004): Pp. 54--87.
- truncNorm, various truncated
normal distributions & random numbers
Utilities
- bin2dec, dec2bin,
[un]compress large binary matrices into a small decimal vectors
- delifAll, selifAll, trimAll, vputAll, delif, selif, trimr, or vput for vectors
of variables
- exist, Does file exist on disk?
- ICPSR, change ICPSR to alphabetical state codes
- in, Tests whether one set is contained in another
- invRV, recode infinities
- ismissM, MKmissM,
element-by-element matrix version of ismiss and miss
- loadA, loadVars
load ascii file into matrix or variables
- makeFac, makeFacN, creates factorial or labeled factorial
designs
- mat2str, str2mat
matrix to string and back
- missR, finds missing values in set of variables
- prtV, nicely print set of variables
- rndInx, randomly permute rows of a set of variables
- scalOne, scalZero, check for scalar one or zero
- seqaS, seqaSE creates
vector of evenly spaced points without or with end points
- strStrip, strip and format a string
- subdat, subdatV
loads submatrix of dataset into matrix or variables
- token2, more flexible string parser,
faster for larger strings
- Vin, checks if variable is in a data buffer
- vreadAll, dumps data buffer to memory in
stored variable names
- vrename, rename element in data buffer
Graphics
- graphGK, reset GK's graphic globals
- graphOn, graphic windows commands for unix
- pLine, pLineS, easily
add lines to graph
- pSym, easily add symbol to graph
- pText, xy with labeled points
- ternary, plot ternary
diagram
- triple, triple scatter plot
(x, y, size of circle)