nathanmct
commented
4 years ago Hi Guillaume,
Thanks so much for the correspondence! I'm glad the paper and code are helpful to you!
Unfortunately I'm not exactly sure what is causing the problem, but I have a series of ideas that you could try: -make sure you have the most recent version of the package installed. Since the 'summary.simmr_output' function wasn't found, it makes me think you have an old or partial version installed. install.packages('simmr') should do the trick. -make sure you have JAGS installed. It runs in the background and does the Bayesian computing. I downloaded it from: https://sourceforge.net/projects/mcmc-jags/files/ . It doesn't need to be open, just installed. -do you have the 'simmrmcmc.r' file in your working directory? This is the function that does the computing in the step "susp_out = simmr_mcmc(simmr_groups_susp)". It sounds like the errors are saying the objects you are trying to plot() and summary() are the wrong length, which could be because they weren't created correctly by the "simmr_mcmc" step. -you can also follow the vignette for the package: https://cran.r-project.org/web/packages/simmr/vignettes/simmr.html The package creator Andrew Parnell has more info on his github site too. -Can you download the data for the Harris et al. paper and run the script on that? That will determine if it's the code that's not working, or if your own input data is not structured correctly. Make sure yours is the exact same as mine. If it's different, say in the number of columns, then you'll have to through the code and make sure the correct columns are called.
Let me know if any of this helps. If it doesn't we can go back to the drawing board.
Best regards, Nathan
On Tue, May 26, 2020 at 11:50 AM GuillaumeBridier notifications@github.com wrote:
Dear Nathan,
Thank you so much to share your R codes based on the analyses performed in Harris et al. 2018. I really liked your paper and your approach on stable isotope analyses. I would like also to make some bayesian mixing models on my own macrofauna data (from NE Greenland) and make a figure similar as your fig.5. I first tried to run our r script with your data. Everything was running well until the part "Visually asses convergence of chains and make note of Gelman diagnostics". Just below this line of code, I faced several issues:
I cannot plot convergence of chains and calculate Gelman diagnostics. R return me this error messages : > plot(susp_out, type = 'convergence')
Error in match.arg(type, several.ok = TRUE) : 'arg' should be one of “isospace”, “histogram”, “density”, “matrix”, “boxplot” > summary(susp_out, type = "diagnostics")
Summary for 1 Error in apply(out_all, 2, "quantile", probs = c(0.025, 0.25, 0.5, 0.75, : dim(X) doit avoir un longueur positive [dim(X) should have a positive length]
I cannot plot OM source correlations: > plot(susp_out, type='matrix', group=1:max(grp.susp))
Error in colnames<-(tmp, value = x$input$source_names) : attempt to set 'colnames' on an object with less than two dimensions
And more importantly, I cannot obtain 95% credible intervals. The summary.simmr_output function was first not find by R (it's a bit weird as I supposed that it should be included inside the simmr package...). But I loaded the code of this function from this page: https://rdrr.io/cran/simmr/src/R/summary.simmr_output.R I then tried to run again the code but R return me this error message: > susp.summ <- summary.simmr_output(susp_out, type=c("quantiles"), group=c(1:max(susp$Code)))
Error in summary.simmr_output(susp_out, type = c("quantiles"), group = c(1:max(susp$Code))) : l'argument 3 correspond à plusieurs arguments formels [arguments 3 corresponds to several formal arguments]
Do you have any idea on what's wrong with my analysis? Thank you very much! :)
All the best, Guillaume
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/nathanmct/Beaufort-Sea-Lagoons/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD2CTMJZN6JQB2UUZO3AL3LRTQMPHANCNFSM4NLATCFQ .
GuillaumeBridier
Dear Nathan,
Thank you so much to share your R codes based on the analyses performed in Harris et al. 2018. I really liked your paper and your approach on stable isotope analyses. I would like also to make some bayesian mixing models on my own macrofauna data (from NE Greenland) and make a figure similar as your fig.5. I first tried to run our r script with your data. Everything was running well until the part "Visually asses convergence of chains and make note of Gelman diagnostics". Just below this line of code, I faced several issues:
I cannot plot convergence of chains and calculate Gelman diagnostics. R return me this error messages : > plot(susp_out, type = 'convergence') Error in match.arg(type, several.ok = TRUE) : 'arg' should be one of “isospace”, “histogram”, “density”, “matrix”, “boxplot” > summary(susp_out, type = "diagnostics") _Summary for 1 Error in apply(outall, 2, "quantile", probs = c(0.025, 0.25, 0.5, 0.75, : dim(X) doit avoir un longueur positive [dim(X) should have a positive length]
I cannot plot OM source correlations: > plot(susp_out, type='matrix', group=1:max(grp.susp)) _Error in
colnames<-(*tmp*, value = x$input$sourcenames) : attempt to set 'colnames' on an object with less than two dimensionsAnd more importantly, I cannot obtain 95% credible intervals. The summary.simmr_output function was first not find by R (it's a bit weird as I supposed that it should be included inside the simmr package...). But I loaded the code of this function from this page: https://rdrr.io/cran/simmr/src/R/summary.simmr_output.R I then tried to run again the code but R return me this error message: > susp.summ <- summary.simmr_output(susp_out, type=c("quantiles"), group=c(1:max(susp$Code))) _Error in summary.simmr_output(suspout, type = c("quantiles"), group = c(1:max(susp$Code))) : l'argument 3 correspond à plusieurs arguments formels [arguments 3 corresponds to several formal arguments]
Do you have any idea on what's wrong with my analysis? Thank you very much! :)
All the best, Guillaume