Tidy a demand-parameter comparison into a flat contrasts frame

Backend-agnostic broom::tidy() method for beezdemand_comparison objects (returned by get_demand_comparisons() on both the NLME and TMB backends). This flat long tibble is the cross-backend contract: identical column names and order regardless of backend. The nested object itself keeps each backend's native dialect (see get_demand_comparisons()).

Usage

# S3 method for class 'beezdemand_comparison'
tidy(x, exponentiate = FALSE, ...)

Arguments

x

A beezdemand_comparison object.

exponentiate

Logical. If TRUE, return base-invariant ratios (estimate = 10^estimate, CIs back-transformed); std.error becomes NA following broom's convention for exponentiated fits. Default FALSE.

...

Unused.

Value

A tibble with columns param, contrast, estimate, std.error, statistic, df, conf.low, conf.high, p.value. Estimates and CIs are on the log10 scale (or ratios when exponentiate = TRUE). statistic is a t ratio with finite df on the NLME backend and an asymptotic z (df = Inf) on the TMB backend (the value differs by backend, by design).

Examples

# \donttest{
data(apt_full)
# 40 subjects per gender keep the example fast; use the full data in practice
ids <- unique(apt_full[c("id", "gender")])
ids <- ids[ids$gender %in% c("Male", "Female"), ]
keep <- unlist(lapply(split(ids$id, ids$gender), head, 40))
dat <- apt_full[apt_full$id %in% keep, ]
fit <- fit_demand_tmb(dat, equation = "exponential",
                      factors = "gender", verbose = 0)
#>   equation='exponential': Dropped 501 zero-consumption observations (859 remaining).
res <- get_demand_comparisons(fit, param = c("Q0", "alpha"))
tidy(res)
#> # A tibble: 2 × 9
#>   param contrast   estimate std.error statistic    df conf.low conf.high p.value
#>   <chr> <chr>         <dbl>     <dbl>     <dbl> <dbl>    <dbl>     <dbl>   <dbl>
#> 1 Q0    Female - …  -0.236     0.0686    -3.45    Inf   -0.371    -0.102 5.69e-4
#> 2 alpha Female - …   0.0358    0.0849     0.422   Inf   -0.131     0.202 6.73e-1
tidy(res, exponentiate = TRUE)
#> # A tibble: 2 × 9
#>   param contrast   estimate std.error statistic    df conf.low conf.high p.value
#>   <chr> <chr>         <dbl>     <dbl>     <dbl> <dbl>    <dbl>     <dbl>   <dbl>
#> 1 Q0    Female - …    0.580        NA    -3.45    Inf    0.426     0.791 5.69e-4
#> 2 alpha Female - …    1.09         NA     0.422   Inf    0.740     1.59  6.73e-1
# }