CODEWITHSUNDEEP
r
Xandian97
Xandian97

Reputation: 51

How can I combine two rows of a dataframe into one?

Basically, I have a simple two column dataframe, one with degrees of a circle in 10° intervals, and the other with the frequencies of those degrees in a different dataframe.

In the degrees column there is a row for 0° and a row for 360°, and since those mean the same thing (in the context of wind direction) I'd like to combine the two rows into a single 0° row. E.g. from:

degree   freq 
0        1446
10       7652
20       2655
...
360      5417

To this

degree   freq 
0        6863
10       7652
20       2655
...

I'm sure this a very simple thing to do but I've spent 3 hours trying to figure it out and have gotten nowhere >.>

Upvotes: 1

Views: 75

Answers (2)

akrun
akrun

Reputation: 886938

One option is to change the value of 360 to 0 and do a group by 'degree' and get the sum of 'freq'

library(dplyr)
df1 %>%
     group_by(degree = replace(degree, degree == 360, 0)) %>%
     summarise(freq = sum(freq))
# A tibble: 3 x 2
#  degree  freq
#*  <dbl> <int>
#1      0  6863
#2     10  7652
#3     20  2655

In base R, we can use aggregate

aggregate(freq ~ degree, transform(df1, 
      degree =  replace(degree, degree == 360, 0)), sum)

Or as @Onyambu commented

aggregate(freq~ degree %% 360,df1,sum)

NOTE: It is a generalized version where multiple elements with 360 can be changed to 0 and then do a group by sum

data

df1 <- structure(list(degree = c(0L, 10L, 20L, 360L), freq = c(1446L, 
7652L, 2655L, 5417L)), class = "data.frame", row.names = c(NA, 
-4L))

Upvotes: 6

MacOS
MacOS

Reputation: 1159

If I read you question correctly, your df has only one row with 360. So, the following may be an option too.

# Add the two rows
df$freq[df$degree == 0] <- 
    df$freq[df$degree == 0] + df$freq[df$degree == 360]

# Delete row
df <- df[ !df$degree==360,]

Upvotes: 1

Related Questions