Design not reproducible despite setting seed

[LucFrancisENX]

Thanks for the package. Despite setting seed, I am getting different results every time I run design.
Many thanks for your help!

[llrs]

Hi thanks for using the package. Could you share some code showing this?
This would help me understand where the problem come from and to check if I have properly fixed it or not.

[LucFrancisENX]

Hi IIrs thanks for your quick response.

Here is some example code

# Set seed 
set.seed(123)
## mock dataframe
df <- data.frame(
  Gender = sample(c("Male", "Female"), 50, replace = TRUE),
  Age = sample(18:60, 50, replace = TRUE),
  Column1 = sample(c(sample(1:100, 40, replace = TRUE), rep(NA, 10)), 50),
  Column2 = sample(c(runif(40, 0, 50), rep(NA, 10)), 50),
  Column3 = sample(c(sample(200:300, 40, replace = TRUE), rep(NA, 10)), 50),
  Column4 = sample(c(rnorm(40, mean = 50, sd = 10), rep(NA, 10)), 50),
  row.names = paste0("ID", 1:50)
)

index_1 <- experDesign::design(df, 13, iterations = 10000)
index_2 <- experDesign::design(df, 13, iterations = 10000)

setequal(index_1, index_2)

#[1] FALSE

[llrs]

Of course, the seed must be set before any call that uses randomization.
Check this:

## mock dataframe
df <- data.frame(
    Gender = sample(c("Male", "Female"), 50, replace = TRUE),
    Age = sample(18:60, 50, replace = TRUE),
    Column1 = sample(c(sample(1:100, 40, replace = TRUE), rep(NA, 10)), 50),
    Column2 = sample(c(runif(40, 0, 50), rep(NA, 10)), 50),
    Column3 = sample(c(sample(200:300, 40, replace = TRUE), rep(NA, 10)), 50),
    Column4 = sample(c(rnorm(40, mean = 50, sd = 10), rep(NA, 10)), 50),
    row.names = paste0("ID", 1:50)
)
set.seed(123)
index_1 <- experDesign::design(df, 13, iterations = 10)
set.seed(123)
index_2 <- experDesign::design(df, 13, iterations = 10)

setequal(index_1, index_2)

Note, I reduced the randomization as we don't really need iterations that long

[LucFrancisENX]

I see, thank you very much!