Search code examples
rdataframedata-manipulation

Convert frequency to raw data values R

Give a dataframe in R

structure(list(Feat1 = c("A", "B", "A", "B"), Time = c("Start", 
"Start", "Finish", "Finish"), Value = c("Value1", "Value2", "Value3", 
"Value2"), feat2 = c("Alpha", "Bravo", "Alpha", "Bravo"), Frequency = c(2L, 
4L, 5L, 3L)), class = "data.frame", row.names = c(NA, -4L))

I have both the frequencies and also features of a certain set of observations. I want all the frequencies to be converted to a number of rows equal to the frequency value while keeping the features.

As an example, the first row becomes these 2 rows because frequency was 2 enter image description here

Solution

  • Try uncount:

    library(tidyr)
dat %>% 
  uncount(Frequency)

    output:

       Feat1   Time  Value feat2
1      A  Start Value1 Alpha
2      A  Start Value1 Alpha
3      B  Start Value2 Bravo
4      B  Start Value2 Bravo
5      B  Start Value2 Bravo
6      B  Start Value2 Bravo
7      A Finish Value3 Alpha
8      A Finish Value3 Alpha
9      A Finish Value3 Alpha
10     A Finish Value3 Alpha
11     A Finish Value3 Alpha
12     B Finish Value2 Bravo
13     B Finish Value2 Bravo
14     B Finish Value2 Bravo