如何在R中的数据框中组合相似的元素

Question

I have a data frame consisting of

Lancaster001A    76
Lancaster001B    35
Lancaster002A    46
Lancaster002D     9
....             ...

I'd like to consolidate the dataframe into this

Lancaster001    111
Lancaster002     55

And so remove the smaller categorising. I couldn't find a way to do with merge, is there a general function that can be used using similarity?

Answer 1

Here is a base R solution using a regex to remove all characters after three numeric characters:

DF <- read.table(text = "Lancaster001A    76
                 Lancaster001B    35
                 Lancaster002A    46
                 Lancaster002D     9")

setNames(aggregate(V2 ~ gsub("(?<=\\d{3}).*", "", V1, perl = TRUE), 
                   DF, FUN = sum), 
         c("V1", "V2"))
#            V1  V2
#1 Lancaster001 111
#2 Lancaster002  55

It would be trivial to use data.table if the aggregation is too slow on a large dataset.

Adjust the regex as needed if the structure of your data is different.

Answer 2

Let's assume these names for your columns, and let's assume the 'smaller categorising' means one letter at the end.

id               value
Lancaster001A    76
Lancaster001B    35
Lancaster002A    46
Lancaster002D     9
....             ...

I use dplyr for everything. Install dplyr , make sure your column names are correct, and then try:

library(dplyr)
mydata %>%
  mutate(id = substr(id, 1, nchar(id)-1) %>% # removes last character
  group_by(id) %>%
  summarize(sum = sum(value))

Answer 3

Edit: An even simpler data.table solution from @Arun's helpful tip:

library(data.table)
dt[, list(sum=sum(value)), by = substr(as.character(id),1,nchar(as.character(id)) - 1)]

             id sum
1: Lancaster001 111
2: Lancaster002  55