Skip to contents

Print and summarize a method comparison

Source: R/accessors-print.R
pvstackr_method_comparison_summary.Rd

print() shows a short overview of a comparison made by pv_compare_methods(). summary() returns the same overview, together with the tables of the comparison, as a list of class summary.pvstackr_method_comparison; its print() method also shows the agreement table.

Usage

# S3 method for class 'pvstackr_method_comparison'
print(x, ...)

# S3 method for class 'pvstackr_method_comparison'
summary(object, ...)

# S3 method for class 'summary.pvstackr_method_comparison'
print(x, ...)

Arguments

...

Ignored.

object, x

A comparison (class pvstackr_method_comparison) from pv_compare_methods(); for the print() method of a summary, the list returned by summary().

Value

print() returns its input invisibly. summary() returns a list of class summary.pvstackr_method_comparison with these elements:

  • reference_method: the label of the reference fit.

  • methods, method_labels and n_methods: the method of each fit (named by label), the labels and the number of fits.

  • n_terms: the number of fixed effects in the comparison.

  • blocked_methods and warning_methods: the labels of the fits with status "blocked" or "warning".

  • interval_note and provenance_note: the texts of the two note lines; interval_note is NA when all intervals are coverage-claimable.

  • estimate_table, diagnostic_table, agreement and timing: the tables described in pv_compare_methods().

  • schema_version, summary_schema_version, source_validation, source_reportability_comparison and validation: the format versions of the comparison and of the summary; a copy of the comparison without its fits and the checksum of that copy; and a SHA-256 checksum of the summary, which print() recomputes to detect changes.

Details

The overview shows the label of the reference fit, the label and the method of each fit (only their number in the summary), the number of fixed effects, and the labels of blocked fits and of fits with status "warning" when there are any. A line beginning "provenance note:" follows for every comparison: it says that agreement is descriptive and that a shared target, combined result or source label is not independent confirmation (pv_compare_methods() explains this). A line beginning "interval note:" is added when some or all intervals are descriptive.

The three methods stop with an error if the comparison or the summary was changed after it was created, and for a comparison saved by pvstackr 0.1.x, which has to be made again from current fits. A summary that contains a stack_psis fit and was saved by an earlier version of pvstackr also stops with an error when printed (see pv_migrate_legacy_psis_fit()).

See also

get_estimates() and get_diagnostics() to read a comparison.