Skip to contents

Estimate MICs and Compare Groups

mic_solve.Rd

High‑level wrapper that returns:

  • delta_mic_results – additive pairwise differences (Δ‑MIC).

  • ratio_mic_results – multiplicative pairwise ratios.

  • dod_ratio_results – difference‑of‑differences on the ratio scale (ratio‑of‑ratios, a classic interaction on the log scale).

  • dod_delta_results – difference‑of‑differences on the additive scale (Δ of Δs).

Usage

mic_solve(
  clm_fit,
  newdata = NULL,
  conc_name,
  transform_fun = log1p,
  inv_transform_fun = expm1,
  alpha = 0.05,
  compare_pairs = "all",
  pvalue_scale = c("lp", "logmic")
)

Arguments

clm_fit

Fitted object from ordinal::clm().

newdata

Data frame with factor combinations to evaluate.

conc_name

Character string giving the raw concentration column.

transform_fun

Transformation used in the model (default log1p).

inv_transform_fun

Inverse transformation (default expm1).

alpha

Confidence‑level significance (default 0.05).

compare_pairs

One of "all" (default) to retain every pairwise comparison, or "share_any" to exclude contrasts where the two groups share no covariate levels in newdata..

pvalue_scale

Which pivot the main P_value uses for pairwise tests: "lp" (difference in lp = log1p(MIC), recommended for calibration) or "logmic" (current Wald on log(MIC) for ratios and MIC scale for deltas).

Value

An object of class "mic_solve" containing the tables above plus mic_estimates.

Examples

if (requireNamespace("ordinal", quietly = TRUE)) {
  ## Toy ordinal dataset
  set.seed(1)
  fit <- ordinal::clm(score ~ strain * treatment + log1p(conc), data = yeast_df)
  res <- mic_solve(fit, conc_name = "conc")
  head(res$ratio_mic_results)
}
#>     Group1   Group2 Ratio_MIC SE_logRatio  CI_Lower  CI_Upper      P_value
#> 1  WT:None Mut:None 0.9977145   0.1778356 0.7040967 1.4137747 9.897342e-01
#> 2  WT:None  WT:Salt 2.2014631   0.1499139 1.6409844 2.9533733 3.220246e-08
#> 3  WT:None Mut:Salt 1.0831448   0.1769435 0.7657235 1.5321494 6.517634e-01
#> 4 Mut:None  WT:Salt 2.2065061   0.1516742 1.6390788 2.9703692 4.157323e-08
#> 5 Mut:None Mut:Salt 1.0856261   0.1785414 0.7650777 1.5404762 6.454120e-01
#> 6  WT:Salt Mut:Salt 0.4920114   0.1505724 0.3662754 0.6609103 8.615838e-07
#>     P_value_lp P_value_WaldMIC
#> 1 9.897342e-01    9.897342e-01
#> 2 3.220246e-08    1.410804e-07
#> 3 6.517634e-01    6.517160e-01
#> 4 4.157323e-08    1.810301e-07
#> 5 6.454120e-01    6.454045e-01
#> 6 8.615838e-07    2.472518e-06