# Calculates the ratios of nominated pairs of predictions stored in an `alldiffs.object`. Ratio predictions and error intervals are formed for two levels of a factor, the `ratio.factor`. For each pair of a level of the `ratio.factor` in `numerator.levels` with a level in `denominator.levels`, the ratio predictions are formed from all combinations of the other factors as the ratio of the two predictions for each combination, along with confidence intervals for the ratio predictions computed using the Fieller (1954) method. The printing of the components produced is controlled by the `tables` argument. ```r ## S3 method for class 'alldiffs' ratioTransform(alldiffs.obj, ratio.factor, numerator.levels, denominator.levels, method = "Fieller", alpha = 0.05, response = NULL, response.title = NULL, tables = "predictions", ...) ``` ## Arguments - `alldiffs.obj`: An `alldiffs.object`. - `ratio.factor`: A `character` string giving the name of the factor for whose levels the ratios are to be calculated. - `numerator.levels`: A `character` string containing the levels of `ratio.factor` to be used as numerators of the ratio. - `denominator.levels`: A `character` string containing the levels of `ratio.factor` to be used as denominators of the ratio. - `method`: A `character` string specifying the method to use in calculating the ratios and their `error.intervals`. At present only `Fieller` is available. For the `Fieller` `method`, ratios of predictions are formed and confidence intervals formed for them using Fieller's (1954) theorem. - `alpha`: A `numeric` giving the significance level for LSDs or one minus the confidence level for confidence intervals. - `response`: A `character` specifying the response variable for the predictions. It is stored as an attribute to the `alldiffs.object` . - `response.title`: A `character` specifying the title for the response variable for the predictions. It is stored as an attribute to the `alldiffs.object`. - `tables`: A `character` vector containing either `none` or `predictions` - ``...``: further arguments passed to `linTransform.alldiffs`. ## Returns A `list` of `predictions.frame`s, each containing the ratio predictions and their confidence limits for a combination of the `numerator.levels` with the `denominator.levels`. It will also contain the values of the variables in the `classify` of `alldiffs.obj` that index the ratio predictions, except that the `ratio.factor` is omitted. If `sortFactor` attribute of the `alldiffs.object` is set and is not the `ratio.factor`, the predictions and their backtransforms will be sorted using the `sortOrder` attribute of the `alldiffs.object`. ## Author(s) Chris Brien ## References Fieller, E. C. (1954). Some Problems in Interval Estimation. **Journal of the Royal Statistical Society.Series B (Methodological)**, 16 , 175-185. ## See Also `pairdiffsTransform`, `linTransform`, `predictPlus.asreml`, `as.alldiffs`, `print.alldiffs`, `sort.alldiffs`, `subset.alldiffs`, `allDifferences.data.frame`, `redoErrorIntervals.alldiffs`, `recalcLSD.alldiffs`, `predictPresent.asreml`, `plotPredictions.data.frame`, `as.Date`, `predict.asreml` ## Examples ```r #### Form the ratios and Fieller CIs for RGR Salinity load(system.file("extdata", "testDiffs.rda", package = "asremlPlus", mustWork = TRUE)) Preds.ratio.RGR <- ratioTransform(diffs.RGR, ratio.factor = "Salinity", numerator.levels = "Salt", denominator.levels = "Control") #### Form the ratios and Fieller CIs for Nitrogen compared to no Nitrogen data("Oats.dat") ## Not run: m1.asr <- asreml(Yield ~ Nitrogen*Variety, random=~Blocks/Wplots, data=Oats.dat) current.asrt <- as.asrtests(m1.asr) wald.tab <- current.asrt$wald.tab Var.diffs <- predictPlus(m1.asr, classify="Nitrogen:Variety", pairwise = TRUE, Vmatrix = TRUE, error.intervals = "halfLeast", LSDtype = "factor", LSDby = "Variety", wald.tab = wald.tab) ## End(Not run) ## Use lme4 and emmmeans to get predictions and associated statistics if (requireNamespace("lmerTest", quietly = TRUE) & requireNamespace("emmeans", quietly = TRUE)) { m1.lmer <- lmerTest::lmer(Yield ~ Nitrogen*Variety + (1|Blocks/Wplots), data=Oats.dat) ## Set up a wald.tab int <- as.data.frame(rbind(rep(NA,4))) rownames(int) <- "(Intercept)" wald.tab <- anova(m1.lmer, ddf = "Kenward", type = 1)[,3:6] names(wald.tab) <- names(int) <- c("Df", "denDF", "F.inc", "Pr") wald.tab <- rbind(int, wald.tab) #Get predictions Var.emm <- emmeans::emmeans(m1.lmer, specs = ~ Nitrogen:Variety) Var.preds <- summary(Var.emm) ## Modify Var.preds to be compatible with a predictions.frame Var.preds <- as.predictions.frame(Var.preds, predictions = "emmean", se = "SE", interval.type = "CI", interval.names = c("lower.CL", "upper.CL")) Var.vcov <- vcov(Var.emm) Var.sed <- NULL den.df <- wald.tab[match("Variety", rownames(wald.tab)), "denDF"] #Create alldiffs object Var.diffs <- as.alldiffs(predictions = Var.preds, sed = Var.sed, vcov = Var.vcov, classify = "Nitrogen:Variety", response = "Yield", tdf = den.df) } if (exists("Var.diffs")) Preds.ratio.OatsN <- ratioTransform(alldiffs.obj = Var.diffs, ratio.factor = "Nitrogen", numerator.levels = c("0.2","0.4","0.6"), denominator.levels = "0.2") ```

ratioTransform.alldiffs function