Splitting column with numeric value into separate rows in R data frame-CodePudding

I have an R dataframe that contains a rather large number of length measurements.

It is structured as follows:

> head(rb.len)
# A tibble: 6 × 5
  Length  Year Value Total Perc_Distr
   <dbl> <dbl> <dbl> <dbl>      <dbl>
1    9.5  1981     1 16641          0
2   10.5  1981     3 16641          0
3   12.5  1981     3 16641          0
4   13.5  1981     4 16641          0
5   14.5  1981    17 16641          0
6   15.5  1981    31 16641          0

Individuals of a certain length are grouped together and the total number of individuals (n) of a length class is listed in the column "value" (e.g. 17 individuals of 14.5cm were measured). For my further analysis I need each measurement to be in a separate row (so basically I need 17 rows with a measurement of 14.5cm). Unfortunately all I have learned so far is how to split columns with observations with multiple delimited values. As I have a single numeric value I am unsure how to proceed.

Hope you can help, Thanks in advance!

CodePudding user response：

1) tidyr Using rb.len shown reproducibly in the Note at the end, use uncount as shown. Add the argument .remove=FALSE to uncount if you prefer to retain the Value column.

library(dplyr)
library(tidyr)

rb.len %>% uncount(Value)
##   Length other
## 1    9.5     a
## 2    9.5     a
## 3   10.5     b
## 4   10.5     b
## 5   10.5     b

2) Base R Using base R we have the following. Replace , ] with , -2] if you prefer to omit Value from the output (because in rb.len shown in the Note Value is the second column).

rb.len[rep(1:nrow(rb.len), rb.len$Value), ]
##     Length Value other
## 1      9.5     2     a
## 1.1    9.5     2     a
## 2     10.5     3     b
## 2.1   10.5     3     b
## 2.2   10.5     3     b

Note

rb.len <- data.frame(Length = c(9.5, 10.5), Value = 2:3, other = letters[1:2])

rb.len
##   Length Value other
## 1    9.5     2     a
## 2   10.5     3     b

CodePudding user response：

I hope I've understood you correctly. There is a nice solution that repeats the whole rows n-times:

rb.len<- as.data.frame(lapply(rb.len, rep, df$Value))

If you wish to multiply only some columns:

rb.len<- as.data.frame(lapply(rb.len[1:2], rep, df$Value))