Thank you for sharing your video! The explanation was very clear! I have a question: Is there a package to run multinomial model with random effects? I see that it isn't possible with nnet.
Thanks for the feedback! Multinomial with random effect? I believe so. But, as soon as I can remember, there was no elegant solution "for all the problems". For example, the first one ist "brms" package. With that you you have to specify the categorical family (multinomial is also possible, but with more complicated structure). As this one is bayesian, you will not get p-values. The second one is "mclogit" package, this one does not allow a quick and nice visualization. But check it out anyway: cran.r-project.org/web/packages/mclogit/mclogit.pdf. Please, let me know what you found as the best solution. cheers
easy ;) library(sjPlot) plot_model(m, type = "pred") check out the function ?plot_model() it does lots of things for bayesian models.... like random effects etc.
I watched this series again and learned a lot more yet again. I wish i could give a second thumbs up. Thanks for the wonderful work.
So nice of you! Watching, commenting and liking is the best support already! So, I thank you and I am glad my content is useful!
Incredible usefull!, thank you very much!, great work!.
You're very welcome, Charlie! I am happy it is useful not only for me! :)
Really appreciate it 🙏 that what I was looking for 😻
Thanks for a nice feedback Youness! :)
Thanks sir very informative. Sir if possible then made video on structure equation modeling...
Thanks for the idea Shah Fahad, I'll put it on the to-do list :)
Thank you for sharing your video! The explanation was very clear! I have a question: Is there a package to run multinomial model with random effects? I see that it isn't possible with nnet.
Thanks for the feedback! Multinomial with random effect? I believe so. But, as soon as I can remember, there was no elegant solution "for all the problems". For example, the first one ist "brms" package. With that you you have to specify the categorical family (multinomial is also possible, but with more complicated structure). As this one is bayesian, you will not get p-values. The second one is "mclogit" package, this one does not allow a quick and nice visualization. But check it out anyway: cran.r-project.org/web/packages/mclogit/mclogit.pdf. Please, let me know what you found as the best solution. cheers
Does the 'performance' package offer diagnostics for Logistic Regression? Can I pass a glm() object to its check_model() function?
Totally! It takes the most models and chooses what assumptions to check. It evolves and is getting better with every update
Hey Yury! How can we overlay exponential, weibull, etc. distributions over Kaplan-Meier Curve in R?
Hey hey, there is a way:
- first install the package:
install.packages("remotes")
remotes::install_github("RichardBirnie/survtools")
- then run the code from this hithub page:
github.com/RichardBirnie/survtools/blob/master/inst/examples/ggsurvplot2_example.R
Hope that helps, cheers!
so i tried the check model function on this script
###Multiple non linear models with interactyion
m
Sure, because alleffects command plots predictions. Use check_model(m) to see all the assumption checks
How about Bayesian statistics and estimate in rstan,rstanarm
easy ;)
library(sjPlot)
plot_model(m, type = "pred")
check out the function ?plot_model() it does lots of things for bayesian models.... like random effects etc.
Is it possible to visualise ordered probit model?
Dear Hoold, yes, I think it is possible. I just quickly tried out an example, and "effects" package plotte it without any problems ;)
dat
Thanks. I tried it but it is not working for me. How can I fix it?
You probably need to install packages first: “foreign” , “MASS”, “effects”