Finding the opposite of condition (using R tidyverse) in order to split dataset in 2

Question

I'm trying to split my dataset (296 rows) in two, the first part will have some conditions and the other is only the rest of the dataset that does not match the conditions.

I did this and found 81 rows for the first part :

cardio = donnees %>%
  select(`Nausées/vomissements`,Vertige,Nystagmus,`Ataxie:Démarche ébrieuse`,`Motif si pas HINTS`,
         Alcool,Tabac,`atcd neuro`,Dyslipidémies,Diabète) %>%
  filter(Alcool == "Yes" |
         Tabac == "Yes"|
         `atcd neuro` == "3" |
         Dyslipidémies == "Yes"|
         Diabète == "Yes")

Then I simply use "!" to find the rest, but unfortunately I only got 77 rows instead of the 215 expected,

donnees %>%
  select(`Nausées/vomissements`,Vertige,Nystagmus,`Ataxie:Démarche ébrieuse`,`Motif si pas HINTS`,
         Alcool,Tabac,`atcd neuro`,Dyslipidémies,Diabète) %>%
  filter(!(Alcool == "Yes" |
           Tabac == "Yes" |
           `atcd neuro` == "3" |
           Dyslipidémies == "Yes" |
           Diabète == "Yes" ))

If someone can help? Thanks a lot

score 2 · Answer 1 · answered Jul 01 '21 at 19:02

I anti_join method suggested by @MonJeanJean should work. But if incase it doesn't, here is a bit different approach, the idea is to create a index column and exclude which ever you don't require (reminiscent of mysql days).

donnees$index = 1:nrow(donnees)

cardio = donnees %>%
  select(`Nausées/vomissements`,Vertige,Nystagmus,`Ataxie:Démarche ébrieuse`,`Motif si pas HINTS`,
         Alcool,Tabac,`atcd neuro`,Dyslipidémies,Diabète, index) %>%
  filter(Alcool == "Yes" |
         Tabac == "Yes"|
         `atcd neuro` == "3" |
         Dyslipidémies == "Yes"|
         Diabète == "Yes")

cardio_required = cardio[-index, ]

This will give you the 215 rows

I thought to this method, but for me it can be too ressources-expensives as soon as you have a lot of row. But it works well for this little sample! — MonJeanJean, Jul 01 '21 at 19:04

MonJeanJean · Accepted Answer · 2021-07-01T18:59:22.263

1

Instead of using the brackets, you can replace the == and | symbols with != and &:

donnees %>%
  select(`Nausées/vomissements`,Vertige,Nystagmus,`Ataxie:Démarche ébrieuse`,`Motif si pas HINTS`,
         Alcool,Tabac,`atcd neuro`,Dyslipidémies,Diabète) %>%
  filter(Alcool != "Yes" &
           Tabac != "Yes" &
           `atcd neuro` != "3" &
           Dyslipidémies != "Yes" &
           Diabète != "Yes" )

Edit: you can use the anti-join function:

cardio = donnees %>%
  select(`Nausées/vomissements`,Vertige,Nystagmus,`Ataxie:Démarche ébrieuse`,`Motif si pas HINTS`,
         Alcool,Tabac,`atcd neuro`,Dyslipidémies,Diabète) %>%
  filter(Alcool == "Yes" |
         Tabac == "Yes"|
         `atcd neuro` == "3" |
         Dyslipidémies == "Yes"|
         Diabète == "Yes")

others <- dplyr::anti_join(donnees, cardio)

edited Jul 01 '21 at 18:59

answered Jul 01 '21 at 18:47

MonJeanJean

2,876
1
4
20

Thanks for the answer, unfortunately I tried this aswell before but the output is 77 rows (exactly the same as my try) – Elysé Barbaza Jul 01 '21 at 18:51
Could you provide a sample using `dput` please? – MonJeanJean Jul 01 '21 at 18:52
Can I show you some screenshots? I do not really know how to use dput ? – Elysé Barbaza Jul 01 '21 at 18:55
@ElyséBarbaza check-out the edit i just made. Plus, here the link to make your example reproductible: https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example – MonJeanJean Jul 01 '21 at 18:59
Thank you a lot of your help it worked, I'm really grateful Have a wonderful evening ! This platform is amazing – Elysé Barbaza Jul 01 '21 at 19:03
Don't forget to validate the answer! Welcome to SO! Enjoy @ElyséBarbaza – MonJeanJean Jul 01 '21 at 19:04

Finding the opposite of condition (using R tidyverse) in order to split dataset in 2

2 Answers2