Rev 3733 | Blame | Compare with Previous | Last modification | View Log | Download | RSS feed
%%% $Id: Nelson.Rd,v 1.4 2003/07/22 19:42:20 bates Exp $\name{Nelson}\alias{Nelson}\non_function{}\title{Dialectric breakdown data}\description{The \code{Nelson} data frame has 128 rows and 3 columns of data from anaccelerated test of dialectric breakdown.}\format{This data frame contains the following columns:\describe{\item{y}{A numeric vector of dialectric breakdown strength values.}\item{x1}{A numeric vector of time values.}\item{x2}{A numeric vector of temperature values.}}}\details{These data are the result of a study involvingthe analysis of performance degradation data fromaccelerated tests, published in IEEE Transactionson Reliability. The response variable is dialectricbreakdown strength in kilo-volts, and the predictorvariables are time in weeks and temperature in degreesCelsius.}\source{Nelson, W. (1981).Analysis of Performance-Degradation Data.IEEE Transactions on Reliability.Vol. 2, R-30, No. 2, pp. 149-155.}\examples{Try <- function(expr) if (!inherits(val <- try(expr), "try-error")) valplot(y ~ x1, data = Nelson, log = "y")plot(y ~ x2, data = Nelson, log = "y")coplot(y ~ x1 | x2, data = Nelson)coplot(y ~ x2 | x1, data = Nelson)Try(fm1 <- nls(log(y) ~ b1 - b2*x1 * exp(-b3*x2), data = Nelson,start = c(b1 = 2, b2 = 0.0001, b3 = -0.01), trace = TRUE))Try(fm1a <- nls(log(y) ~ b1 - b2*x1 * exp(-b3*x2), data = Nelson,trace = TRUE, alg = "port",start = c(b1 = 2, b2 = 0.0001, b3 = -0.01)))Try(fm2 <- nls(log(y) ~ b1 - b2*x1 * exp(-b3*x2), data = Nelson,start = c(b1 = 2.5, b2 = 0.000000005, b3 = -0.05), trace = TRUE))Try(fm2 <- nls(log(y) ~ b1 - b2*x1 * exp(-b3*x2), data = Nelson,trace = TRUE, alg = "port",start = c(b1 = 2.5, b2 = 0.000000005, b3 = -0.05)))Try(fm3 <- nls(log(y) ~ cbind(1, -x1 * exp(-b3*x2)), data = Nelson,start = c(b3 = -0.01), trace = TRUE, algorithm = "plinear"))Try(fm4 <- nls(log(y) ~ cbind(1, -x1 * exp(-b3*x2)), data = Nelson,start = c(b3 = -0.05), trace = TRUE, algorithm = "plinear"))}\keyword{datasets}