spdata {QRMlib} | R Documentation |
The spdata
timeSeries dataset has 100 rows and 3 columns.
It contains default data for A, BBB, BB, B and C-rated companies for the years 1981 to 2000
data(spdata)
a matrix containing 100 rows and 4 columns.
The colums are:
rating | rating category (A, BBB, BB, B, CCC) |
firms | number of companies in rating category |
number of defaults | number of companies defaulting in category |
The rows are the years from 1981-2000
documentation by Scott Ulman for R-language distribution
Standard and Poors Credit Monitor
#Must attach MASS and nlme libraries to run mixed effect regression model library(MASS); library(nlme); #Load timeSeries: data(spdata); #timeSeries ratingval <- spdata@recordIDs$rating; yearval <- as.numeric(spdata@recordIDs$DATE); #Use R- library MASS to get glmmPQL which runs a mixed-effects model. #It will measure random effects and fixed effects. #'year' -'ratings' determine the unique results(20 years 1981-2000 with 5 obligor #class ratings each year) results <- glmmPQL(cbind(defaults,firms-defaults) ~ -1 + ratingval, random = ~1| yearval, family=binomial(probit), data=spdata); results; summary(results); summary(results)$tTable[,1]; detach("package:nlme"); detach("package:MASS");