Wsrtjtyk

I have a long format data frame that looks like this:

What I'd like to do is (1) randomly split the trials from Condition 1 into 4 groups and then calculate the mean of Y for each ID and (2) carry out this same procedure for the trials in Condition 2. This is what this new output would look like:

Is there a concise way to do this using the tidyverse? I'm just getting started and am having a having a hard time with this!

Assuming that your data frame is not that small, this should do exactly what you need.

set.seed(2)

df <- tibble(ID = c(rep(1,50),rep(2,50)),

       Condition = rep(c(1,1,2,2,2,1,1,2,2,2,1,1,1,2,2,1,1,1,2,2),5),

       Y = rnorm(100,100,20)) ##Creates some fake data







result_df <- df %>% mutate(randomizer = sapply(1:length(row_number()), function(i)sample(c(1,2,3,4),1))) %>%

  group_by(Condition) %>%

  group_by(ID, Condition, randomizer) %>%

  summarize(mean_y = mean(Y, na.rm = TRUE)) %>%

  mutate(condition_status = paste0("Condition",Condition,"M",randomizer)) %>%

  ungroup() %>% 

  dplyr::select(-Condition, -randomizer) %>%

  spread(condition_status, mean_y)





result_df

Not the best way but it does work

x <- data.frame(c(rep(1,10), rep(2,10)), c(1,1,2,2,2,1,1,2,2,2,1,1,1,2,2,1,1,1,2,2), c(100,200,80,58,89,100,200,80,58,89,95,72,99,120,130,95,72,99,120,130))

colnames(x) <- c("ID", "Condition", "Y")

id <- unique(x[,1])

con <- unique(x[,2])

#multiple by the number of groups to split into

y <- data.frame(matrix(, ncol = (length(id) * 4) + 1, nrow = length(id)))

y[,1] <- id

colnames(y) <- c("ID", "Condition1M1", "Condition1M2", "Condition1M3", "Condition1M4", "Condition2M1", "Condition2M2", "Condition2M3", "Condition2M4")

x1 <- x[which(x[,2] == con[1]),]

x2 <- x[which(x[,2] == con[2]),]

#specify sample size

n <- 5

x11 <- x1[sample(nrow(x1), n, replace = FALSE),]

x12 <- x1[sample(nrow(x1), n, replace = FALSE),]

x13 <- x1[sample(nrow(x1), n, replace = FALSE),]

x14 <- x1[sample(nrow(x1), n, replace = FALSE),]

x21 <- x2[sample(nrow(x2), n, replace = FALSE),]

x22 <- x2[sample(nrow(x2), n, replace = FALSE),]

x23 <- x2[sample(nrow(x2), n, replace = FALSE),]

x24 <- x2[sample(nrow(x2), n, replace = FALSE),]



y[1,2] <- mean(x11[which(x11[,1] == id[1]),3])

y[1,3] <- mean(x12[which(x12[,1] == id[1]),3])

y[1,4] <- mean(x13[which(x13[,1] == id[1]),3])

y[1,5] <- mean(x14[which(x14[,1] == id[1]),3])

y[2,2] <- mean(x21[which(x21[,1] == id[2]),3])

y[2,3] <- mean(x22[which(x22[,1] == id[2]),3])

y[2,4] <- mean(x23[which(x23[,1] == id[2]),3])

y[2,5] <- mean(x24[which(x24[,1] == id[2]),3])

y[1,6] <- mean(x11[which(x11[,1] == id[1]),3])

y[1,7] <- mean(x12[which(x12[,1] == id[1]),3])

y[1,8] <- mean(x13[which(x13[,1] == id[1]),3])

y[1,9] <- mean(x14[which(x14[,1] == id[1]),3])

y[2,6] <- mean(x21[which(x21[,1] == id[2]),3])

y[2,7] <- mean(x22[which(x22[,1] == id[2]),3])

y[2,8] <- mean(x23[which(x23[,1] == id[2]),3])

y[2,9] <- mean(x24[which(x24[,1] == id[2]),3])

Results:

  ID Condition1M1 Condition1M2 Condition1M3 Condition1M4 Condition2M1 Condition2M2 Condition2M3 Condition2M4

1  1     133.3333          100          150     133.3333     133.3333          100          150     133.3333

2  2     125.0000          125          125     120.0000     125.0000          125          125     120.0000