CODEWITHSUNDEEP
r
Pree
Pree

Reputation: 77

Random number generation in R with multiple conditions

I have this problem below to address. Your inputs/pointers are very much appreciated.Thanks in advance!

I have a data.table called Defect. And the first column is Length

Defect Table : 

Length

6355
6236
1345 
3123

I want to add two more columns called DefectStarts and DefectEnds. I would like to use random function to determine where the defect starts and where it ends.

Basic criteria is as follows:

 DefectStarts and DefectEnds <= Length

 DefectEnds > DefectStarts

 DefectEnds - DefectStarts <= Length

Final data should something like this(values could be different as long as it satisfies above conditions)

 Defect table:

Length     DefectStarts     DefectEnds

6355         1234             4356

6236         2000             4567

1345         500               689

3123         342               3120

Upvotes: 1

Views: 1524

Answers (4)

Ajay Ohri
Ajay Ohri

Reputation: 3492

Psuedocode:

i for (1 to length(df){
  df[i,DefectStarts]=sample(df[i,Length],1,T)
  df[i,DefectEnds]=sample(df[i,Length]-df[i,DefectStarts],1,T)
}

Upvotes: 0

Hong Ooi
Hong Ooi

Reputation: 57686

runif(n, min, max) generates n random numbers with a range from min to max. The latter two arguments can be vectors.

n <- nrow(df)
df$start <- runif(n, max=df$Length)
df$end <- runif(n, min=df$start, max=df$Length)

Using the dplyr package:

df <- mutate(df, start=runif(n(), max=Length), end=runif(n(), min=start, max=Length))

Upvotes: 3

Pree
Pree

Reputation: 77

After some messing around with the code, this seemed to do the job :

Defect <- Defect[,DefectStarts:=runif(row_count,min=1,max=Defect$Length)]

Defect <- Defect[,DefectEnds:=runif(row_count,min=Defect$DefectStarts,max=Defect$Length)]

Thanks again for your support! Much appreciated. And please do suggest if there are any better ways to do this.

Upvotes: 0

amonk
amonk

Reputation: 1795

Consider df to be your dataframe:

 df$random<-runif(nrow(df)) #just create a random column and fill it with random numbers

apply the condition

 df$random<-ifelse(df$DefectEnds > df$DefectStarts,df$random,NA)#select rows with this condition

Upvotes: 0

Related Questions