Package index • fleetcheck

The claims register

read_claims(): Read the claims register
claims_summary(): Counts by verdict
scoreboard(): The scoreboard, as printable text
claims_list_md(): The register as a numbered list
scoreboard_md(): The register as a markdown table
badges_md(): The README badge row
verdict_html(): A verdict as its HTML lozenge
check_claims(): Fail if the register has regressed

Comparison statistics

Every statistic quoted anywhere in this project. Defined once and unit-tested, because the alternative is what happened to fleet’s own documentation: the same quantity derived twice, in two articles, wrong both times.

agreement(): Agreement between two models on the same quantity
BAND_K BURDEN_MIN: How a replicate band and a burden floor are defined
replicate_band(): The band a set of replicates produces
band_z(): Standardised departure from a set of replicates
outside_rates(): How often a candidate falls outside the replicate band, against the IBM
band_position() inside_band(): Where a value sits relative to a replicate band
band_summary(): Summarise a set of band comparisons as a criterion outcome

Provenance and paths

stamp(): Stamp an output with what produced it
round_sig(): Round the numeric columns of a data frame
fc_root(): The repository root
fc_results(): Where a validation keeps its committed results
find_claims(): Locate the claims register
replace_block(): Replace a marked block in a file
md_cell(): Make a string safe inside a markdown table cell

Scenario constants

The settings the comparison scripts share. Exported so the scripts run against an installed package and not only under pkgload::load_all().

VDIR() BURN_Y POP N_REP AGE_EDGES SEASON EIR_GRID EIR_REF PROFILE_EIR INT_LABELS TS_LABELS TS_YEARS TS_NET_EVERY: Shared scenario constants
int_scenario() int_parts(): Intervention scenario names across the transmission grid