如何基于R中的另一列创建具有多个值的新列

Question

I have a data frame in R called A.Data.

It has 8 different columns: plate , row, col , TOF , EXT , green , red , and yellow .

Below is an example of the data.

> head(A.Data)
    plate row col TOF EXT green red yellow 
1     1   A  12  20  21     2   0      0      
2     1   C  12  20  17     0   1      0      
3     1   C  11  20  17     0   0      1      
4     1   A  10  20  16     1   1      3      
5     1   A  10  20  16     0   0      0      
6     1   A  10  20  15     0   0      0

I'm trying to add a new column to A.Data called conc (short for concentration). The new column called conc depends on the value in the col column.

-If col is 1 or 7, conc should equal to 0
-If col is 2 or 8, conc should equal to 0.5
-If col is 3 or 9, conc should equal to 1
-If col is 4 or 10, conc should equal to 2
-If col is 5 or 11, conc should equal to 4
-If col is 6 or 12, conc should say NA

So for the first 6 rows of data, the conc column should say NA, NA, 4, 2, 2, 2 because the col column values for the first 6 rows are 12, 12, 11, 10, 10, 10 .

I asked my professor for help and he gave me this hint:

df$newcol <- rep(1, 1000) will add a new column to the df data frame called newcol and will have 1 replicated 1000 times

Try to add a concentration column called conc with 0, 0.5, 1, 2, 4, NA replicated as many times as you need for the entire column.

Here is the summary of A.Data$col , in case you might find it useful...

> summary (A.Data$col)
   1    2    3    4    5    6    7    8    9   10   11   12 NA's 
1128  703  538  256  156   30 2101 1039  741  294   73   60   11

Thank you!

Answer 1

Not tested, but this may work

map_column <- rep(c(0, 0.5, 1, 2, 4, NA),2)
df$newcol <- map_column[df$col]

EDIT: The idea behind this code is: map_column , which is a vector of length 12, serves here as a map (in the mathematical sense) between the numbers 1 to 12 and the values in the vector. For instance,

map_column[[1]]

returns the first element of the vector (0), and

map_column[[9]]

returns the 9th element of the vector (1), and so on. Now R vectors have the capability to process several inputs at once, so that

map_column[c(1,9)]

returns the corresponding elements ( c(0,1) ) at these positions in one go. Note that it is important to use a single square bracket [ instead of [[ here.

Answer 2

This works.

convert <- function(number){
  if(number == 1 | number == 7){return(0)}
  if(number == 2 | number == 8){return(.5)}
  if(number == 3 | number == 9){return(1)}
  if(number == 4 | number == 10){return(2)}
  if(number == 5 | number == 11){return(4)}
  if(number == 6 | number == 12){return(NA)}
}

A.Data$newcol <- do.call(rbind, lapply(A.Data$col, convert))

Answer 3

Use merge.

augment <- data.frame(col=1:12,conc=rep(c(0, 0.5, 1, 2, 4, NA),2))
A.Data  <-merge(A.Data,augment,by="col",sort=F)
A.Data
#   col plate row TOF EXT green red yellow conc
# 1  12     1   A  20  21     2   0      0   NA
# 2  12     1   C  20  17     0   1      0   NA
# 3  11     1   C  20  17     0   0      1    4
# 4  10     1   A  20  16     1   1      3    2
# 5  10     1   A  20  16     0   0      0    2
# 6  10     1   A  20  15     0   0      0    2

This creates an an augment dataframe with 2 columns, col corresponding to col in A.Data , and conc with the augment. Then merge that with A.Data based on col.

Answer 4

Here's a very different approach based on mathematical and logical operations:

x <- c(1:12, NA) # an example vector including all possible values

floor(2 ^ (z <- x %% 6 - 2)) / 2 * (z + 2 | NA)

The result:

[1] 0.0 0.5 1.0 2.0 4.0  NA 0.0 0.5 1.0 2.0 4.0  NA  NA

(I fear that this solution may appear like obfuscation.)