rethinking2020
is a study project of Statistical Rethinking, second edition, by Richard McElreath. [@elreath2020]
Many thanks to Richard McElreath who wrote Statistical Rethinking to make Bayesian statistics so interesting.
This study project is inspired by Solomon Kurtz who did wonderful work. It can be found at Solomon Kurtz.
The packages used in this project are as follows.
Package | Comment |
---|---|
rethinking | This R package accompanies a course and book on Bayesian data analysis: McElreath 2020. Statistical Rethinking, 2nd edition. |
brms | Used to do pretty much all models and computations. |
loo | Used by both rethinking and brms to compute WAIC and LOO. See section 7.5 in chapter for more details. |
posterior | Used to work with the output of Bayesian models. |
tidybayes | Bayesian statistics do require a lot of data processing. tidybayes is incredibly helpful in providing a constant and efficient way of coding the data processing. |
bayestestR | A comprehensive and consistent set of functions to analyze and describe posterior distributions. |
tidyverse | The package for data science in general. |
extraDistr | has many extra distributions such as BetaBinom, NegBetaBinom, HalfCauchy, ZIP, Pareto, etc. which are useful in real-world applications and usually dispersed among many different packages. |
ggmcmc | tool for assessing and diagnosing convergence of Markov Chain Monte Carlo simulations, as well as for graphically display results from full MCMC analysis |
simpr | tidyverse-friendly framework for generating simulated data, fitting models on simulations, and tidying model results. |
modelr | functions that help you create elegant pipelines when modelling |
gtools | functions to assist in R programming, including |
logit and inv.logit functions used in many places in this project |
Package | Comment |
---|---|
ggdist | tidybayes relies heavily on the ggdist and reexport very many of it's functions. In addition ggdist is useful for both a frequencist as well as a bayesian approach. |
bayesplot | Used as complement when ggdist and tidybayes do not provide a satisfactory solution. |
ggplot2 | Vizualisations using the grammar of graphics. |
scales | Provide the internal scaling infrastructure used by ggplot2 |
paletteer | A nice package to manage color and shape palettes. |
Package | Comment |
---|---|
dagitty | Draw and analyse causal diagrams |
ggdag | R package to visualize and analyze causal directed acyclic graphs |
tidygraph | To manipulate graph data |
ggraph | To visualize graphs |
Package | Comment |
---|---|
xfun | The function xfun::cache_rds is used for caching which is used several times in almost every chapter. |