dsalfran / imputerobust Goto Github PK
View Code? Open in Web Editor NEWMultiple Imputation with GAMLSS
Multiple Imputation with GAMLSS
Hello,
I'm trying to impute missing data from a highly overdispersed and zero-inflated count variable and came across your package.
After failing to use the 'gamlssZIP' on my data I tried to run the example from the help page and it seems to fail too (although there are no error messages), as there are no imputed values for X.2 and X.3 but only NAs when I do imputed.sets$imp$X.2
.
I asked a colleague to try to run the example and it worked on his computer, which had the same package versions loaded but just differed by the fact that it's 64-bit Windows while mine is 32-bit.
I then asked him to run the function on a subset of my data but he got the same issue of having 'NA' instead of imputed values.
Below is the code I used and my sessionInfo()
(I'm not attaching the data on GitHub, but can maybe send a subset privately if I get the authorization).
Thanks in advance for your help!
Thomas
sub_hydrat <- read.table("Stats_Hydratation/sub_data_Thomas.csv", header = TRUE, sep = ";")
head(sub_hydrat)
apply(sub_hydrat, 2, function(x){any(is.na(x))})
only GPAQ_ptotal has missing values
init_im <- mice(sub_hydrat, maxit=0, print=F)
im_pred <- init_im$predictorMatrix
im_pred[c("date_inclu", "FUP_irt", "irt", "FUP_deces", "deces", "conso_eau", "conso_eau_cls", "SEXE", "region_new", "AGEI", "origine_fac", "saison_fac", "tabac_fac", "neph_type_fac", "stadeincb", "fluide_ss_eau", "dfgin"),] <- 0
ZI_meth <- init_im$method
ZI_meth["GPAQ_ptotal"] <- "gamlssZIP"
im_ZIP <- mice(sub_hydrat, m = 1, maxit = 2, seed = 300119, meth = ZI_meth, pred = im_pred, n.cyc = 1, bf.cyc = 1, cyc = 1)
im_ZIP$imp$GPAQ_ptotal
gives a vector of NA instead of imputed data ...`
R version 3.5.1 (2018-07-02)
Platform: i386-w64-mingw32/i386 (32-bit)
Running under: Windows 7 (build 7601) Service Pack 1
Matrix products: default
locale:
[1] LC_COLLATE=French_France.1252 LC_CTYPE=French_France.1252
[3] LC_MONETARY=French_France.1252 LC_NUMERIC=C
[5] LC_TIME=French_France.1252
attached base packages:
[1] parallel splines stats graphics grDevices utils
[7] datasets methods base
other attached packages:
[1] bindrcpp_0.2.2 ImputeRobust_1.3-1 mice_3.3.0
[4] lattice_0.20-35 gamlss_5.1-2 nlme_3.1-137
[7] gamlss.dist_5.1-1 MASS_7.3-50 gamlss.data_5.1-0
loaded via a namespace (and not attached):
[1] digest_0.6.18 minqa_1.2.4 R6_2.3.0
[4] gWidgets_0.0-54.1 assertthat_0.2.0 grid_3.5.1
[7] tcltk_3.5.1 gWidgetstcltk_0.0-55 broom_0.5.1
[10] survival_2.42-3 tidyselect_0.2.5 extremevalues_2.3.2
[13] generics_0.0.2 lme4_1.1-19 pillar_1.3.0
[16] compiler_3.5.1 tibble_1.4.2 tidyr_0.8.2
[19] nnet_7.3-12 pkgconfig_2.0.2 pan_1.6
[22] purrr_0.2.5 Matrix_1.2-15 rstudioapi_0.8
[25] glue_1.3.0 mitml_0.3-7 magrittr_1.5
[28] rlang_0.3.0.1 yaml_2.2.0 tools_3.5.1
[31] bindr_0.1.1 dplyr_0.7.8 jomo_2.6-6
[34] nloptr_1.2.1 crayon_1.3.4 rpart_4.1-13
[37] backports_1.1.2 Rcpp_1.0.0
Hello,
Thanks for contributing this package to CRAN. I have been looking for a way to impute data for variance function regression for quite a while.
I have noticed what I think are potentially bugs in the package. I am using version 1.1-1 and mice 2.30.
When running the example provided (the one in the documentation under the topic mice.impute.gamlss, I am getting the following error:
In file(file, if (append) "a" else "w") :
cannot open file '/dev/null': No such file or directory
and even though the program finishes and produces an object, there are no imputed values, and all missing values are still NA.
This error repeats several times when the code runs (12 times for the example above).
Thanks!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.