Skip to contents

Trim activation columns for all result files in a directory.

Source: R/cleaning.R
trim_directory.Rd

Applies trim_results to every buzzdetect result file found recursively in dir_results, saving the trimmed output to path_out.

Usage

trim_directory(
  dir_results,
  path_out,
  activation_digits,
  neurons_keep = NULL,
  output_format = "rds",
  if_exists = "stop",
  workers = 1
)

Arguments

dir_results

Path to the directory containing buzzdetect result files.

path_out

Path to the output file (.rds or .csv) or output directory. Directories are created automatically if they do not exist. When a file path is given, its extension overrides output_format.

activation_digits

Integer. Number of decimal places to round activation values to.

neurons_keep

Character vector of neuron names to retain (see trim_results). If NULL (default), all neurons are kept.

output_format

Output file format when path_out is a directory. One of "rds" (default) or "csv". Ignored when path_out is a file path.

if_exists

What to do if an output file already exists. One of "stop" (default, throws an error), "skip" (silently skips existing files), or "overwrite" (overwrites with a warning).

workers

Number of parallel workers. Defaults to 1 (sequential). Parallelism uses parallel::mcmapply and may not be supported on all platforms. Ignored when path_out is a single file.

Value

Invisibly returns the output file path(s).

Details

If path_out ends in .rds or .csv, all results are read via read_directory (with return_ident = TRUE), trimmed, and written to that single file (overriding output_format). Otherwise, path_out is treated as an output directory and files are written there preserving the original directory structure.

See also

trim_results for the single-file version.

Examples

if (FALSE) { # \dontrun{
dir_in  <- system.file('extdata/five_flowers', package = 'buzzr')
dir_out <- file.path(tempdir(), 'five_flowers_trimmed')

# Trim all files, rounding to 2 decimal places, saved as .rds in a directory
trim_directory(dir_in, dir_out, activation_digits = 2)

# Combine all results into a single .rds file
trim_directory(dir_in, file.path(tempdir(), 'trimmed.rds'), activation_digits = 2)

# Combine all results into a single CSV
trim_directory(dir_in, file.path(tempdir(), 'trimmed.csv'), activation_digits = 2)

# Re-run, keeping only the ins_buzz neuron and overwriting existing files
trim_directory(
  dir_in, dir_out,
  activation_digits = 2,
  neurons_keep = 'ins_buzz',
  if_exists = 'overwrite'
)
} # }