mattfidler
commented
5 years ago Hi @ONYLAB,
The RxODE::rxProgressAbort() in this case means that the more robust linear standard error calculation was aborted, and it falls back to the approximated fisher information matrix for the covariance calculation. In fact, the fisher information matrix doesn't do a great job here either because it is corrected to be symmetric, so I wouldn't trust the SEs very much.
The warnings show these conditions:
1: In (function (uif, data, est = NULL, control = list(), ..., sum.prod = FALSE, :
Covariance matrix non-positive definite, corrected by sqrtm(fim %*% fim)
2: In (function (uif, data, est = NULL, control = list(), ..., sum.prod = FALSE, :
Linearization of FIM could not be used to calculate covariance.Perhaps the abort could be more informative.
ONYLAB
Hello nlmixr developers,
A very new user here... I would like to use an ode model to fit my data using nlmixr saem.
With
nlmixr(vivo, dat, est="saem"), R crashes after 149th iterationWith
fit <- nlmixr(vivo, dat, est="saem", saemControl(n.burn=90, n.em=50)), I receive the following messages, but nlmixr moves on with calculating residuals/tables:Calculating covariance matrix [====|====|====|====|====|====|====|====|====|====] 100%; 0.054 sec Error in RxODE::rxProgressAbort() : Aborted calculation Load into sympy...Using sympy via reticulate doneWhat does the Error in RxODE::rxProgressAbort() mean here exactly? Could there be something wrong with my model syntax?
Thank you!
My code is here and I used a small dataset.
My sessioninfo() output is: R version 3.6.0 (2019-04-26) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows 7 x64 (build 7601) Service Pack 1
Matrix products: default
Random number generation: RNG: Mersenne-Twister Normal: Inversion Sample: Rounding
locale: [1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252
attached base packages: [1] stats graphics grDevices utils datasets methods base
other attached packages: [1] xpose.nlmixr_0.1.4 xpose_0.4.4 ggplot2_3.1.1 nlmixr_1.1.0-9
[5] RxODE_0.9.0-6
loaded via a namespace (and not attached): [1] reticulate_1.12 tidyselect_0.2.5 xfun_0.7
[4] purrr_0.3.2 lattice_0.20-38 colorspace_1.4-1
[7] generics_0.0.2 vpc_1.1.0 rlang_0.3.4
[10] huxtable_4.5.0 pillar_1.4.1 glue_1.3.1
[13] withr_2.1.2 tweenr_1.0.1 plyr_1.8.4
[16] stringr_1.4.0 munsell_0.5.0 n1qn1_6.0.1-3
[19] gtable_0.3.0 mvnfast_0.2.5 memoise_1.1.0
[22] labeling_0.3 knitr_1.23 RcppArmadillo_0.9.400.3.0 [25] rex_1.1.2 parallel_3.6.0 sys_3.2
[28] Rcpp_1.0.1 scales_1.0.0 jsonlite_1.6
[31] farver_1.1.0 brew_1.0-6 ggforce_0.2.2
[34] digest_0.6.19 lotri_0.1.1 stringi_1.4.3
[37] dplyr_0.8.1 lbfgs_1.2.1 polyclip_1.10-0
[40] grid_3.6.0 tools_3.6.0 magrittr_1.5
[43] dparser_0.1.8 lazyeval_0.2.2 tibble_2.1.2
[46] PreciseSums_0.3 crayon_1.3.4 tidyr_0.8.3
[49] pkgconfig_2.0.2 Matrix_1.2-17 MASS_7.3-51.4
[52] assertthat_0.2.1 rstudioapi_0.10 R6_2.4.0
[55] units_0.6-3 nlme_3.1-139 compiler_3.6.0