This repository contains metadata and functions to create spatial null models for FreeSurfer surface data. Many of these functions are wrappers for functions provided by the BrainSMASH and neuromaps toolboxes.
The package relies on a working installation of python, we well as the python toolboxes BrainSMASH and neuromaps. These need to be installed prior to the fsnulls package. Also, you have to initiate python in R using the reticulate package.
These can be specified in the .Rprofile file, e.g.,
if (benchmarkme::get_cpu()$model_name == "Apple M1 Max") {
Sys.setenv(RETICULATE_PYTHON = "/Users/sphache/miniconda3/bin/python")
} else {
Sys.setenv(RETICULATE_PYTHON = "/opt/anaconda3/bin/python")
}
library(reticulate)
reticulate::py_config()
or
python.path <- "/opt/anaconda3/bin/python"
reticulate::use_python(python.path)Given the large amount of metadata that is contained within the package, it uses git annex to store the data on a remove repository other than github. You therefore also need to install (datalad)[https://www.datalad.org], which can be used to clone the repository and to download the data.
To clone the repository, use
datalad clone https://github.com/christineecker/fsnulls.git
To download the data, use
datalad get data/
datalad get inst/extdata
# etc.