In a large dataframe ("bind") with three columns I have to add a fourth column with values conditionally based on the first three columns.

It's about travel. A person (id) has several options to travel (e.g. bus, car, train, ...), that is always quite different. Sometimes a person has several options by train to travel from A to B for example.

My dataframe looks like this:

Now I want to create a new column "train_avail". If the train journey takes twice as long as the car journey, then I would like to write train_avail = 0 in a new column otherwise write 1.

What i would like to have:

This is, what I tried, but obviously doesn't work. In my complete cluelessness, I have no idea how to write the code.

bind %>%
  group_by(id) %>%
  mutate(train_avail = ifelse((bind$main_mode = train) & (duration = 2 * (bind$main_mode = car)) ~ '0', '1'))

I'm sorry, I have no idea at all how this should work. I've only been working with R for a few days.

Thanks for your help!!

CodePudding user response：

df |>
group_by(id) |>
  mutate(train_avail = {
    car <- min(duration[main_mode == "car"])
    if_else(main_mode == "train",
            if_else(duration > 2* car, 0, 1),
            NA_real_)
    })

##>   # A tibble: 8 × 4
##> # Groups:   id [3]
##>   main_mode duration    id train_avail
##>   <chr>        <int> <int>       <dbl>
##> 1 train         1250     1           0
##> 2 train          900     1           1
##> 3 car            540     1          NA
##> 4 train          650     2           1
##> 5 car            450     2          NA
##> 6 plane          350     2          NA
##> 7 train          350     3           1
##> 8 car            890     3          NA