CODEWITHSUNDEEP
rdplyrgroup-by
Skyk
Skyk

Reputation: 27

How to use the group_by function while keeping other columns

I wanted to know how it is possible to use the group_by() function with being able to keep other columns in the process.

My dataset consists of a lot of different information about different birds and I need to calculate frequencies of 7 behaviors for each day for each bird.

Let's say I subset my dataset for one individual:

alm18subset <- subset(df, df$individualID == "Almen18 (eobs 5861)")

I then have the data for only this individual.

I then want to know the frequencies of the behavior for each day in a new dataframe using group_by:

alm18freq <- alm18subset %>%
  group_by(agesincetaggingdays, behaviors) %>%
  summarise(n = n()) %>%
  mutate(freq = n / sum(n))

which gives me a output of age, behaviors, and then summarise the number of time each behavior is carried out each day and the frequency I'm looking for. My question is, in this process I lose all the other columns using the group_by(). I want to keep some information in my data, like sex or the time a behavior was carried out. This is because I need to do models.

Any idea of how to do it ?

Then I also need to filter according to the behavior:

alm18active <- filter(alm18freq, rf8fitted == "Active")

Thank you !

Upvotes: 0

Views: 94

Answers (2)

Andre Wildberg
Andre Wildberg

Reputation: 19211

Replacing summarize with mutate should suffice in your case. The subsequent filter can be done right after this operation.

alm18freq_active <- alm18subset %>%
  group_by(agesincetaggingdays, behaviors) %>%
  mutate(n = n(), freq = n / sum(n)) %>%
  filter(rf8fitted == "Active")

Upvotes: 1

Gero
Gero

Reputation: 177

You can simply merge the two dataframes that you have by their common columns, this will include the frequency column in the initial subset:

merge(alm18subset, alm18freq)

Upvotes: 1

Related Questions