CODEWITHSUNDEEP
r
user315648
user315648

Reputation: 1965

Finding nearest value in a list

I have a list of percentiles and I wish to query a nearest value from it by specifying a probability and vice versa. Consider the following code:

samples <- runif(1000, 0, 1)
percentiles <- quantile(samples,probs=(0:100)/100)
plot(percentiles)

I can easily read the plot(e.g. 50% prob -> 0.5 in this example), however how can I find the nearest value in the list (either by specifying the probability or the value on the Y-axis) ?

Thanks!

Upvotes: 0

Views: 97

Answers (2)

Rich Scriven
Rich Scriven

Reputation: 99331

Building off Robert's answer, you can look at all the neighbors with this. Coincidentally, the result is in list form.

> zzz <- sapply(1:(length(percentiles)-1), function(i){
    c(sort(percentiles)[i-1], sort(percentiles)[i+1])
    }, USE.NAMES = TRUE)[-1]

The use of USE.NAMES allows you to access the neighbors easily via the list index.

> zzz[99]  ## neighbors of the 99th percentile
##       98%      100% 
## 0.9861953 0.9986512 
> zzz[25]  ## neighbors of the 25th percentile
##       24%       26% 
## 0.2654937 0.2926444 
> zzz[36]  ## neighbors of the 36th percentile
##       35%       37% 
## 0.3756176 0.3864981

Upvotes: 2

Robert Krzyzanowski
Robert Krzyzanowski

Reputation: 9344

Sort first and look at the neighbors.

i <- 50
sorted_percentiles <- sort(percentiles) 
# the values adjacent to the 50th data point are:
sorted_percentiles[i - 1]
sorted_percentiles[i + 1]

Upvotes: 2

Related Questions