I am looking for a way to use the values in the first row to help filter values I want. Say if I want to keep certain columns in R based on the values in the first row. So in the first row, we have -0.5, 0.7, 1.1, and -1.2.

I want to only keep values that are equal to or greater than 1, or less than or equal to -1.2. Everything else will just be dropped.

So say my original data I have is DF1

With the filter, I would now have a new DF:

What would be the best way for this. I feel there is a way to sort columns based on values from a row, but I am unable to think of the way we can with certain commands.

ID <- c("First", "Second", "Second", "Third", "Third", "Fourth")
Location <- c("Park", "School", "Home", "Car", "Lake", "Prison")
XPL <- c(-0.5,2,4,1,7,4)
SNA <- c(0.7,5,5,8,5,5)
AAS <- c(1.1,2,6,8,4,1)
APA <- c(-1.2,3,4,5,6,7)
DF1 <- data.frame(ID, Location, XPL, SNA, AAS,APA)

CodePudding user response：

If you are using the first row as the basis, you can convert it to a normal integer vector and use the which function to know the indexes that will be kept.

test.row <- as.numeric(DF1[1,3:6])

the 3 and 6 corresponds to the range of index from XPL to APA.

DF1 <- DF1[,c(1:2, 2   which(test.row >= 1 | test.row <= -1.2))]

we keep the columns ID and Location as 1:2 and we offset 2 to the which function.

CodePudding user response：

In dplyr, you can select numeric columns whose first absolute value is above 1:

library(dplyr)
DF1 %>% 
  select(!where(~ is.numeric(.x) && abs(first(.x)) <= 1))

#       ID Location AAS  APA GOP
# 1  First     Park 1.1 -1.2 1.4
# 2 Second   School 2.0  3.0 1.0
# 3 Second     Home 6.0  4.0 2.0
# 4  Third      Car 8.0  5.0 2.0
# 5  Third     Lake 4.0  6.0 3.0
# 6 Fourth   Prison 1.0  7.0 3.0

Or with between:

DF1 %>% 
  select(!where(~ is.numeric(.x) && between(first(.x), -1.19, 0.99)))