check_grain: Determine the grain of each resistance outcome in timber and...

In iAM-AMR/sawmill: Process Queries from the CEDAR Database

Determine the grain of each resistance outcome in timber and validate the grains' fields

Description

The check_grain() function checks that a resistance outcome format is specified, determines its grain and calls functions to validate the grains' fields. See details for more information.

Usage

check_grain(timber)

Arguments

timber

A tibble of timber.

Details

The odds ratio (and its confidence intervals) are the universal measure of CEDAR and sawmill. A resistance outcome may be specified using the odds ratio directly, or by using a two-by-two table of counts or prevalences. sawmill supports four resistance outcome formats:

• Contingency table

• Prevalence table

• Odds ratio

• Risk ratio

Resistance Outcome Formats

There are two types of two-by-two tables supported by sawmill: contingency tables (containing count data), and prevalence tables (containing prevalence (

Contingency Table

	AMR+	AMR-	Total
Exposed	A	B	M1
Referent	C	D	M2

Prevalence Table

	AMR+	AMR-	Total
Exposed	P%	R%	M1
Referent	Q%	S%	M2

There are two types of measures of association supported by sawmill: odds ratios and risk ratios.

Odds Ratio

	Lower CI	Odds Ratio	Upper CI
	odds_lo	odds	odds_up

Risk Ratio

	Lower CI	Risk Ratio	Upper CI
	rratio_lo	rratio	rratio_up

Supported Grains

A 'grain' is a set of fields from which we can derive an odds ratio and its confidence intervals. For tabular data, supported grains are a set of fields from which we can first compute a complete table of counts (A, B, C, D), then compute an odds ratio and its confidence intervals.

There are currently seven supported grains:

• con_table_pos_neg: A, B, C, D

• con_table_pos_tot: A, C, M1, M2

• con_table_neg_tot: A, C, M1, M2

• prev_table_pos_tot: P, Q, M1, M2

• prev_table_neg_tot: R, S, M1, M2

• odds_ratio: odds_ratio_lo, odds_ratio, odds_ratio_up

• risk_ratio: risk_ratio_lo, risk_ratio, risk_ratio_up

Unsupported Grains

An 'unsupported grain' is a set of fields from which we cannot derive an odds ratio and it's confidence intervals. These include grains where any requisite fields are NA.

The following grains are explicitly unsupported:

• prev_table_pos_neg: P, Q, R, S

sawmill fields

TODO: Link to documentation on sawmill fields.

Value

A tibble of timber with sawmill status fields and additional field grain. Where checks fail, grain is NA.

iAM-AMR/sawmill documentation built on June 30, 2024, 2:25 a.m.