CODEWITHSUNDEEP
rdata.table
Fpertille
Fpertille

Reputation: 383

Creating a condition to generate a column from other two columns

How can I create a column in a dataset conditioned to other two columns according to this example.

On this example I am showing all possible combinations from my data:

library(data.table)
data <- "chr start tag depth BCV
    chr1 3273 chr1-3273 0 0
    chr1 3274 chr1-3274 1 1 
    chr1 3275 chr1-3275 1 2
    chr1 3276 chr1-3276 1 3
    chr1 3277 chr1-3277 2 1
    chr1 3278 chr1-3278 2 2
    chr1 3279 chr1-3279 2 3
    chr1 3280 chr1-3280 3 1
    chr1 3281 chr1-3281 3 2
    chr1 3282 chr1-3282 3 3"
    data <- read.table(text=data, header=T)

Expected outcome:

newdata <- "chr start tag depth BCV states
chr1 3273 chr1-3273 0 0 0
chr1 3274 chr1-3274 1 1 1
chr1 3275 chr1-3275 1 2 2
chr1 3276 chr1-3276 1 3 3
chr1 3277 chr1-3277 2 1 4
chr1 3278 chr1-3278 2 2 5
chr1 3279 chr1-3279 2 3 6
chr1 3280 chr1-3280 3 1 7
chr1 3281 chr1-3281 3 2 8
chr1 3282 chr1-3282 3 3 9" 
newdata <- read.table(text=newdata, header=T)

The requested column in the data.table (states) ranges from 0 to 9 depending on the combination of the other two columns.

Upvotes: 0

Views: 64

Answers (3)

jogo
jogo

Reputation: 12559

this gives you the desired table:

setDT(data)
data[, states:=.I-1]
data

To ensure the order of your records with respect to the two columns depthand BCV you can set keys in the datatable before the data[, states:=.I-1].

Upvotes: 0

Gregor Thomas
Gregor Thomas

Reputation: 145755

In base R,

data$states = as.numeric(factor(paste(data$depth, data$BCV))) - 1

Upvotes: 1

eddi
eddi

Reputation: 49448

OP is not well phrased. I'm guessing your looking for:

setDT(data)[, states := .GRP - 1, by = .(depth, BCV)]

Upvotes: 3

Related Questions