Re-run a recorded fit

Rebuild the fit function from its recipe in metadata(se)$provenance and fit the data again. The rows of fit_source in the output assay (and in the summary assay, if any) are replaced; other rows are kept.

Usage

refit_from_provenance(se, assay, fit_source, allow_source = FALSE)

Arguments

se

SummarizedExperiment

assay

string; name of the assay holding the fit results.

fit_source

string; value of the fit_source column.

allow_source

logical; evaluate fit functions stored as source code.

Value

se with the fit re-run.

Details

A function from an installed package is taken from that package. A warning is raised when the installed version or the code differs from the recorded one, since results may then differ.

A function stored as source code is evaluated only with allow_source = TRUE: the object may come from someone else, and evaluating its code runs that code in your session.

See also

check_reproducibility

Examples

mae <- get_synthetic_data("finalMAE_small.qs2")
se <- mae[["single-agent"]]
se <- apply_fit(
  se, fit_drug_response_metrics, "single-agent",
  output_assay = "custom_hill", fit_source = "hill"
)
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
se <- refit_from_provenance(se, "custom_hill", "hill")
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced
#> Warning: NaNs produced