Contents

schemaComparison

Association schema for formal model/effect/group comparisons ("comparison")

Column and artifact contract for results of formal model, effect and group comparisons.

Usage

schemaComparison()

Value

A PolyGeniusSchema for the "comparison" analysis type.

Columns

Column	Status	Type	Meaning
`.fit`	required	integer	Integer key joining the row to the per-fit `$fits` lookup.
`fit.id`	required	character	Identifier of the fit the row came from.
`schema`	required	character	Name of the schema the row follows.
`family`	required	character	Fitted regression or comparison family.
`outcome`	required	character	Outcome label.
`predictor`	required	character	Predictor label.
`interaction`	required	character
`covariates`	required	character	Covariates adjusted for.
`term`	required	character	Coefficient term the row describes.
`term.type`	required	character	Kind of coefficient term.
`stratum`	required	character	Stratum the fit was run in.
`formula`	required	character	Model formula of the fit.
`estimate`	required	numeric	Point estimate, on the scale given by `effect.scale`.
`se`	required	numeric	Standard error of the estimate.
`lower`	required	numeric	Lower interval bound.
`upper`	required	numeric	Upper interval bound.
`statistic`	required	numeric	Test statistic.
`pval`	required	numeric	Raw p-value.
`adj.pval`	required	numeric	p-value adjusted within its declared adjustment family.
`adj.family.id`	required	integer	Key into `$indices$multiplicity`, the adjustment family `adj.pval` came from.
`n`	required	numeric	Sample size.
`effect.scale`	required	character	Scale of `estimate`, one of the controlled `effect.scale` vocabulary.
`test`	required	character
`comparison.type`	optional	character
`groups.by`	optional	character
`contrast`	optional	character
`base.model`	optional	character
`full.model`	optional	character
`outcome.type`	optional	character
`n.events`	optional	numeric	Number of events.
`df1`	optional	numeric
`df2`	optional	numeric

Artifacts

Artifact	Status	Federation-safe	Meaning
`observed.curves`	optional	yes	Observed curves per time point and group.
`predicted.curves`	optional	yes	Predicted curves per time point and profile.

Allowed plots: forest (default forest).

Meta-analysis

Not meta-eligible.