Adding Top Level Function to use calculateDelta in factorial designs (WIP)

[lboel]

A small proposal for a top level function to use calculateDelta in a factorial design.

dat <- createDF(nSubjects = 1, nTrl = 10000, design = list("Comp" = c("comp", "incomp")))

dat <- addDataDF(dat,
                  RT = list("Comp_comp"   = c(500, 80, 100),
                           "Comp_incomp" = c(600, 80, 140)))

dat$additionalFactor <- sample(c("Test","Experiment"),nrow(dat),replace = T)

delta <- calculateDeltaFactorialLong(dat, factorNames = c("additionalFactor"),
                                     nDelta = 19,
                                     tDelta = 1,
                                     columns = c("Subject", "Comp", "RT"),
                                     compCoding = c("comp", "incomp"),
                                     quantileType = 5)`

Would output something like this:

This function was developed based on an idea by https://github.com/rubenellinghaus.

At the moment this is only a draft and the main contributors should decide whether I should proceed here and this is a good addition to the package.

Best,

Luke

[lboel]

@igmmgi Ruben uses this approach for his experiment designs. So I build a function based on his approach.

[igmmgi]

Would something like this work using group_modify from dplyr? This keeps the same tibble columns and just adds in the additionalFactor.

library(dplyr)
dat <- createDF(nSubjects = 1, nTrl = 10000, design = list("Comp" = c("comp", "incomp")))

dat <- addDataDF(dat,
                 RT = list("Comp_comp"   = c(500, 80, 100),
                           "Comp_incomp" = c(600, 80, 140)))

dat$additionalFactor <- sample(c("Test","Experiment"),nrow(dat),replace = T)

delta_group <- dat %>%
  group_by(additionalFactor) %>%
  group_modify(~ calculateDelta(.,  
                             nDelta = 19, 
                             tDelta = 1,
                             columns = c("Subject", "Comp", "RT"),
                             compCoding = c("comp", "incomp"),
                             quantileType = 5))

[rubenellinghaus]

Ians solution works well for me. Just adding

%>%
dplyr::group_by(Bin, additionalFactor) %>%
dplyr::summarise(mean_Bin = mean(meanBin), mean_effect = mean(Effect))

gives the tibble I want for ggplot.

Thanks a lot, @igmmgi, this will save me a lot of time and nerves. I was wondering whether the solution might be good to add to the documentation/tutorial, as users without strong programming background might not come up with such straightforward solution. This is just an intuition and up to the developers of course.