It's good you've added references to the (new) help page for varCompProp(). transit time shows up as a kind of random noise within the device, and this is directly proportional to the frequency of operation.
The attached patch adds a variance function structure varConstPropĪlong with example code, a test using nlme and stratified values for t1 and t2 as well as an updated NAMESPACE file and updated documentation.Īpplied to the current revision of R-packages/trunk/nlme (r7890), it passes R CMD check -as-cran with the current development version of R (r79355) as well as R CMD check with the released version R 4.0.3, both on Linux. Where v denotes the variance covariate, s2(v) denotes the variance function evaluated at v, and t, t1 and t2 are the variance function coefficients. model with an additive unobserved effect. fixed effects (typical values) and the structural model (covariate relationship with age, weight, height etc). In the analysis of heteroscedastic longitudonal data, an assumption that is commonly encountered says that there is an additive and a multiplicative or proportional component to the residual variance.Ĭurrently, in nlme we have the following related possibilities varExp