variables named `df` in the globalenv() cannot be used/executed in the multiverse?

[abhsarma]

This is currently throwing an error:

M = multiverse()
df <- data.frame(x = 1:10) %>% mutate( y = x^2 + sample(10:20, 10))
M$df <- ~ df %>%
      mutate( z = branch( value_y, 
                          "none" ~ y,
                          "log" ~ log(y)
                        ))

This throws an error as well:

M = multiverse()
df <- data.frame(x = 1:10) %>% mutate( y = x^2 + sample(10:20, 10))
M$df.s <- ~ df %>%
      mutate( z = branch( value_y, 
                          "none" ~ y,
                          "log" ~ log(y)
                        ))

However, this does not throw an error:

M = multiverse()
data.raw <- data.frame(x = 1:10) %>% mutate( y = x^2 + sample(10:20, 10))
M$df <- ~ data.raw %>%
      mutate( z = branch( value_y, 
                          "none" ~ y,
                          "log" ~ log(y)
                        ))

[abhsarma]

@mjskay I think this is assigning the stats function df here?

If so, solving #18 should address this as well?

[mjskay]

Yeah if I replace the initial .results environments in the multiverse table with the global environment it seems to work. So I would say those environments probably need to be adjusted to be child environments of the environment that first calls multiverse()

[abhsarma]

I am not able to replicate this error currently, so I assume that we'd solved this at some point

[mjskay]

Have you added a test for it? If not it would be a good idea so it doesn't come back

[abhsarma]

Sounds good