CODEWITHSUNDEEP
r
oercim
oercim

Reputation: 1848

Filtering a data frame according to value's specific ranks

This is an updated version of the problem:

Filtering a data frame according to values rank

I have a data frame(df) such as

group  value  x
a      4.2    2
a      4.5    2
a      6.2    2
b      5.1    3
b      3.5    3
a      4.2    2
a      5.1    2
b      6.4    3
b      3.3    3
b      4.1    3
a      5.0    2

The desired output is

group  value
a      4.5
a      6.2  
a      5.1
a      5.0
b      5.1
b      6.4

Namely, x assigns a value to each group.

desired output extracts the

The desired output includes all rows of df except related rows to these values. How can I do that with R? I will be vey glad for any help. Thanks a lot.

Upvotes: 0

Views: 80

Answers (2)

David Arenburg
David Arenburg

Reputation: 92282

Here's another possible data.table approach without reordering the data

library(data.table)
setDT(df)[, value[rank(value, ties.method = "first") > x[1L]], group]
#    group  V1
# 1:     a 4.5
# 2:     a 6.2
# 3:     a 5.1
# 4:     a 5.0
# 5:     b 5.1
# 6:     b 6.4

Or some type of base R approach

df$indx <- with(df, ave(value, group, FUN = rank, ties.method = "first"))
do.call(rbind, lapply(split(df, df$group), function(y) y[y$indx > unique(y$x), ]))
#      group value x indx
# a.2      a   4.5 2    3
# a.3      a   6.2 2    6
# a.7      a   5.1 2    5
# a.11     a   5.0 2    4
# b.4      b   5.1 3    4
# b.8      b   6.4 3    5

Upvotes: 3

akrun
akrun

Reputation: 886988

You could try

library(dplyr)
df1 %>% 
    group_by(group) %>% 
    arrange(value) %>% 
    slice(-(1:x[1]))%>%
    select(-x)
#    group value
#1     a   4.5
#2     a   5.0
#3     a   5.1
#4     a   6.2
#5     b   5.1
#6     b   6.4

Or using base R

 df2 <- df1[order(df1$group, df1$value),]
 indx <-  !!with(df2, ave(x, x, FUN=function(x) c(rep(0,x[1]), 
                         rep(1, length(x)-x[1]))))
 subset(df2, indx, select=-x)

Upvotes: 2

Related Questions