Twitter Spread R-package

R-package with the code used for the paper: Modelling rapid online cultural transmission: Evaluating neutral models on Twitter data with Approximate Bayesian Computation in Palgrave Communication special issue on Cultural Evolution. Vignettes are availalbe online:

• spreadrt.html
• abc_spreadrt.html

Install the package

To install the 'spreadrt' package directly from this git repository you can use different option:

1. Clone the repository

git clone https://github.com/simoncarrignon/spreadrt.git

and install from source:

install.packages("spreadrt",repos=NULL,type="source")
library(spreadrt)

2. Or you can use install_github function from package devtools:

library(devtools)
devtools::install_github("simoncarrignon/spreadrt")

Start using the package

Once you installed the package, the best way to start learning and using the package is to use the vignettes available in vignettes/. You can read directly the .Rmd files and run directly the chunk of code from here or you can compile the code using the library

library(devtools)

and the function:

build_vignettes()

warning: Building the vignettes takes time (more than 30 minutes) as it has to run a great number of simulations and generate numerous graphs.

Once it's done, you can open the resulting html files that should have been created in inst/doc/ at the follwing URLs: spreadrt.html and abc_spreadrt.html.

A simpler and more straightforward way is to go to the compiled version of the vignettes in the pages branch of the github:

• spreadrt.html
• abc_spreadrt.html

Modify the package

If you plan to modify the package and want to check your modifications you need to clone the repository:

git clone https://github.com/simoncarrignon/spreadrt.git

Then start `R` from the git folder and use the packages `devtools` and `roxygen2`
```R
library(devtools)
library(roxygen2)

then if you modify some code within the package you use:

• load_all() to update the code
• documentation() to update the documentation

Approximate Bayesian Computation and exec subfolder

Some of the scripts used to run the ABC have been stored in exec/abcdir as well as some script used to generate the plot of the paper in exec/palcomm. Though those script are very hardware and problem specific they could be of some use (for people using SLURM and MPI mainly).

TODO: Part of script and comment in exec should be integrated in functions and vignette in the main package.