Skip to contents

Save a complete MRMhubExperiment as an RDS file

Source: R/data-export.R
save_dataset_rds.Rd

Writes the whole MRMhubExperiment to a single .rds file: raw data, metadata joins, calibration fits, quantification results and processing history all travel together. This makes the .rds a self-contained, portable snapshot of a complete analysis project — a collaborator can open it with readRDS() to inspect, or load mrmhub to re-plot and re-process, without reconstructing the original file paths, input files or metadata versions.

By default a content fingerprint of the object (rlang::hash()) is embedded in the saved file and printed. Record it (or compare printed fingerprints) to confirm that a reloaded file — or a copy shared with a colleague — holds the identical object. read_dataset_rds() reads the file back and verifies the fingerprint.

Usage

save_dataset_rds(
  data = NULL,
  path,
  hash = TRUE,
  compress = TRUE,
  overwrite = TRUE,
  create_dir = TRUE
)

Arguments

data

A MRMhubExperiment object to save.

path

A character string with the file path. If it does not end in .rds, the extension is appended automatically.

hash

A logical value. If TRUE (default), a content fingerprint of the object is computed with rlang::hash(), embedded in the saved file and printed. The fingerprint is stored as an object attribute and is stripped again by read_dataset_rds(), so it does not affect the reloaded object.

compress

A logical value passed to saveRDS(). If TRUE (the default), the file is gzip-compressed (typically about ten times smaller); FALSE writes it uncompressed (faster, but much larger). Either way the file is read back identically — the compression is detected automatically on load.

overwrite

A logical value indicating whether to overwrite the file if it already exists. Default is TRUE.

create_dir

A logical value. If TRUE (the default), the parent directory of path is created if it does not yet exist.

Value

The path written to, invisibly.

Archiving

The file is a plain R serialization, so it can always be reopened with base R's readRDS() — mrmhub is not required just to read it. Every slot (the dataset, the annot_* metadata tables, and the QC and calibration metrics) is stored as an ordinary tibble and can be inspected via the @ slots or attributes(), which makes the .rds a convenient self-contained archive of a complete analysis. mrmhub (a compatible version) is only needed to re-plot or re-process the object through its methods. For long-term, cross-tool preservation, complement the .rds with an open-standard export (save_dataset_mztab(), save_dataset_summarizedexperiment() or save_dataset_csv()).

See also

read_dataset_rds(), save_report_xlsx(), save_dataset_summarizedexperiment()

Examples

mexp <- data_load_example()
#> ✔ Loaded example dataset 1: 499 analyses and 29 features.
path <- file.path(tempdir(), "example_mexp.rds")
save_dataset_rds(mexp, path)
#> ✔ MRMhubExperiment saved to /tmp/Rtmp7Sqs4e/example_mexp.rds.
#> Content fingerprint: "0e81ef3de5a90827c3ed651abfa25c2d"