Script for downloading data from osf

get_osf_data

# when data are in private repo on osf, we want to download them prior to the analysis

# load tidyverse
library(tidyverse)

# install package to communicate with osf.io. Check the github repo for guidelines
library(osfr)

# function to download data using dplyr do() function
download_files <- function(df, local_data_pth, should_overwrite = T) {
  
  # we need to set correct class as the current version of osfr does not works with dplyr properly
  class(df) <- c("osf_tbl_file","osf_tbl", class(df)) 
  
  df %>% 
    rowwise() %>% 
    do(osf_retrieve_file(.$id) %>% 
         osf_download(path = file.path(local_data_pth, .$name), 
                      overwrite = should_overwrite))
}

# set local data path
local_data_pth <- file.path("data")

# You need to add your PAT key into separate file. This line is used for authentication
osf_auth(token = read_lines("osf_token_write_fd.txt"))

# add guid of your repo
data_guid <- "XXX"
  
  
osf_project <- osf_retrieve_node(data_guid)

# create local structure for data
if(!dir.exists(local_data_pth)) {
  dir.create(local_data_pth,recursive = T) 
}

# select the files that are nested in osf structure
data_files <- 
    osf_project %>% 
    osf_ls_files() %>% 
    filter(name == "directory1") %>%
    osf_ls_files() %>% 
    filter(name == "directory2") %>% 
    osf_ls_files()
    
    
# and download files locally  
data_files %>% 
  do(download_files(.,local_data_pth))