In the domain of clinical research, where the stakes are as high as the complexities of the data, a new statistical aid emerges: bayer: https://github.com/cccnrc/bayer
This R package is not just an advancement in analytics - it’s a revolution in how researchers can approach data, infer significance, and derive conclusions
What Makes `Bayer` Stand Out?
At its heart, bayer is about making Bayesian analysis robust yet accessible. Born from the powerful synergy with the wonderful brms::brm() function, it simplifies the complex, making the potent Bayesian methods a tool for every researcher’s arsenal.
Streamlined Workflow
bayer offers a seamless experience, from model specification to result interpretation, ensuring that researchers can focus on the science, not the syntax.
Rich Visual Insights
Understanding the impact of variables is no longer a trudge through tables. bayer brings you rich visualizations, like the one above, providing a clear and intuitive understanding of posterior distributions and trace plots.
Big Insights
Clinical trials, especially in rare diseases, often grapple with small sample sizes. `Bayer` rises to the challenge, effectively leveraging prior knowledge to bring out the significance that other methods miss.
Prior Knowledge as a Pillar
Every study builds on the shoulders of giants. `Bayer` respects this, allowing the integration of existing expertise and findings to refine models and enhance the precision of predictions.
From Zero to Bayesian Hero
The bayer package ensures that installation and application are as straightforward as possible. With just a few lines of R code, you’re on your way from data to decision:
# Installation devtools::install_github(“cccnrc/bayer”)# Example Usage: Bayesian Logistic Regression library(bayer) model_logistic
Perhaps you misspecified this as discussion and not an article or result? best
Hi Wim, this is not an arrivle yet, I started a discussion to give visibility to the package here in RG too, in order to have people trying it and reporting bugs etc before publication
Enrico Cocchi many thanks for sharing the package with us. I just had a brief look and tried it out, but I wondered how to formulate more complex models.
For example, I have 2 or more time points and I want to predict sigma as well. How can I formulate a repeated t-test like model, which allows for heterogenity? I saw that you implemented student_t as robust alternative to the gaussian likelihood, which is nice!
The example is quite straight foward in brms:
brm(data =data,
family = student(),
formula = bf(outcome ~ 0 + Intercept + time + (1 | ID),
sigma ~ 0 + Intercept + time)
)
Is this possible in bayer? Are there other options available to plot the data or to manipulate the posterior? With the indicator variable approach as above, I need to calculate the mu for time2 from the posterior of Intercept and b_time. Can this be done within bayer? How about (standardized) effect sizes?
Many thanks in advance for answering and taking your time!
Hello Rainer Duesing,
Thank you for reaching out with your suggestions. I appreciate your interest in enhancing the bayer package. Currently, the released version doesn't include options for specifying sigma and mu. However, I'm more than happy to adding these features.
To make sure the enhancements meet your needs and are clear to all users, including those who may not be as familiar with Bayesian terminology, could you provide additional details? A brief explanation of these parameters would be beneficial, both for implementation purposes and for creating user-friendly documentation.
Could you please submit an enhancement request on our GitHub issues page here: https://github.com/cccnrc/bayer/issues/new ?
It would be great if you could outline your specific use case and include the introduction to "sigma" and "mu" as mentioned. This will allow us to collaborate effectively and ensure we don’t overflow this conversation thread.
As example, I see that if I use mtcars and mpg as outcome and hp as time, create a model with formula:
brm( data = mtcars,
family = student(),
formula = bf( mpg ~ 0 + Intercept + hp, sigma ~ 0 + Intercept + hp) )
I can extract:
- b_Intercept
- b_hp
- b_sigma_Intercept
- b_sigma_hp
- nu
from the posterior draws. How would you like to analyze and plot them?
Looking forward to your contribution and working together to improve bayer on github: https://github.com/cccnrc/bayer/issues/new
Hi, thanks for your reply.
The model does look fine as an example. What really would help to improve the workflow would be to :
1) get a back transformed metric for sigma for example. As it is, brms would estimate the sigma intercept and b on a log scale to prevent values
Hi Rainer Duesing
If you can wrap this up in a Github issue (enhancement label: https://github.com/cccnrc/bayer/issues/new ) it would be great as it allows for more straightforward code etc.
If you know what you need I can help in automate it inside the package, but I need some assistance to be sure to actually understand what you need as you seem much more expert on Bayesian theoretic than I am. This would greatly benefit all users!
Thanks again
Hi Rainer Duesing I just pushed a new version with enhancement you suggested, please have a look and let me know! https://github.com/cccnrc/bayer/issues/1#issuecomment-2093388192
Hi Enrico!
Many thanks for your efforts!!! I will try it out as soon as possible and will provide feedback on github!
All the best,
Rainer
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India