Chapter 2 Warning message: In sqrt(-2) : NaNs produced Chapter 3 Warning messages: 1: In cor.test.default(H$G, H$FL, method = "spearman") : Cannot compute exact p-values with ties 2: In cor.test.default(H$G, H$FL, method = "spearman") : Cannot compute exact p-values with ties 3: In boot.ci(tau.bs, conf = 0.95) : bootstrap variances needed for studentized intervals Comments: The method defaults to exact, but because of ties, uses Normal approximation. The boostrap for studentized intervals requires that the use also sample the variance statistic. Normally a statistic is generated (e.g. say mean(y) but also var(y) may also be collected, then the output is a vector of length two (mean,var). Comments: Chapter 4 Warning message: In sample(2500, m, replace = TRUE, prob = prob) : Walker's alias method used: results are different from R < 2.2.0 Ignorable warning message. Chapter 6 Warning message: In sub(object$syntax$docexpr, val, chunk[pos[1L]]) : argument 'replacement' has length > 1 and only the first element will be used val has length > 1 not sure of context. It is a warning message from Sweave... Chapter 8 Warning messages: 1: In rq.fit.br(x, y, tau = tau, ci = TRUE, ...) : Solution may be nonunique 2: In sqrt(diag(z$cov)) : NaNs produced 3: In sqrt(diag(z$cov)) : NaNs produced 4: In sqrt(diag(z$cov)) : NaNs produced In rq.fit.br, means that there are not enough points in simplex, more than one solution is feasible for the bounds. Occurs in the call to rq.fit.br, when the lower and upper bonds are requested. The inversion of ranks has multiple solutions. It can be ignored, but not removed. The other three are from a gpd fit and it occurs when there is not enough data to fit the model. Can be ignored (should be obvious from return level plots). Cannot be removed. Chapter 12 Warning messages: 1: glm.fit: fitted rates numerically 0 occurred 2: glm.fit: algorithm did not converge 3: glm.fit: fitted rates numerically 0 occurred 4: glm.fit: algorithm did not converge 5: glm.fit: fitted rates numerically 0 occurred 6: glm.fit: fitted rates numerically 0 occurred The model has a fitted value near zero in a Poisson model. Need to determine if the fit is used (which I suspect it is not). Warnings cannot be removed, unless the fit is not attempted. Chapter 13 Warning messages: 1: In get.tracks(x = best.use, locations = loc, umin = 64, N = 200) : We found only 48 tracks 2: In dgamma(x, shape, scale, log) : NaNs produced 3: In dgamma(x, shape, scale, log) : NaNs produced 4: In dgamma(x, shape, scale, log) : NaNs produced 5: In dgamma(x, shape, scale, log) : NaNs produced 6: In rq.fit.br(x, y, tau = tau, ...) : Solution may be nonunique 7: In rq.fit.br(x, y, tau = tau, ...) : Solution may be nonunique We only found 48 tracks, more an informational, we asked for 200. Change N to 48 would remove warning. The dgamma function calculated with shape<0. In the rq function, the error occurs in the calculation of the lower and upper bounds, the solution is find, but the bounds are not unique. The last two messages are not serious, however they cannot be removed easily.