unlist is handy and you should use it

Y’all should be using unlist(). unlist is a crazy handy function when you pair it with the purrr library. It’ll take a list and try to give you a vector. Strictly speaking it’s probably best to be using map_dbl() or map_chr() instead but let’s not worry about that at the moment.

Why do I love unlist so much? Because unlist(map( )) gives you a flexible, effective way to iterate (that is parallel friendly with minor changes).

Here’s an example: I have a bunch of traces in a folder and I need to know if there are any experiments that didn’t copy. Additionally, I’d like to know what kind of data is in the file. I could look through them one by one, but that’s no fun. Thankfully, R has functions that perform similarly to shell functions. (More on these some other time.)

So we start by defining a data.frame with all file names and their sizes. Note that calling unlist(map()) inside data.frame() lets us do a lot work very quickly. We find all the files, get information about each one, and then selectively return the size.

traces_dir <- "C:/Users/Daniel/Documents/Trace_Holding"

files_df <- data.frame(files = list.files(traces_dir),
                       bytes = unlist(map(list.files(traces_dir), 
                                          function(abf){ <http://file.info|file.info>(paste0(traces_dir, "/",abf))$size })))

#              files    bytes
# 1 190808a_0000.abf 15006208 <- long gap free recording
# 2 190808a_0001.abf 15006208
# 3 190808a_0002.abf 15006208

The experiment is embedded in the file name for each. Once again , with a little help from unlist we can split the file names into a list of list (i.e. "190808a_0000.abf" becomes [[1]] [[1]] "190808a" [[2]] "0000.abf" select only the first part and populate a new column.

files_df$Experiment <- unlist(transpose(str_split(files_df$files, pattern = "_"))[[1]])

Okay, now we can make use of this. I’ve defined a data.frame for metadata about the experiments.

file_groups
#    Experiment       Group
# 1      190924    Baseline
# ... 
# 19     190918     Delayed

We can join these data frames and apply a little tidyverse magic to see what experiments are missing (we could also compare the sets of experiments directly).

full_join(file_groups, files_df) %>% 
  filter(<http://is.na|is.na>(files)) %>% 
  group_by(Group, Experiment) %>% 
  tally()

#  Experiments that didn't transfer:

#   Group       Experiment     n
# 1 Baseline    190924         1
# 2 Baseline    190924a        1
# ...

We can repeat the same strategy to programattically look at the protocol types (e.g. based on file size or channel number via <http://file.info|file.info> | readABF()). Moral of the story, you should so stop applying yourself and give unlist and purrr functions a try.