Dplyr select throws a resolve to integer error, but basic select works, why?

Question

I have a large spreaded data frame:

df: a1 a2 a3 a4 a5 ...............
    r  w  sd w  y ........

I have another input which is a subset of df.

subset_df: a3 a4 a5
           f  e  u

My goal is to take the column names of subset_df, select these columns in df and continue from there (in my case to compare the values).

When I do this the simple way:

df[,names(subset_df)] it works, but why it refuses to work with dplyr select?

Here is the error when running:

names_sub_df <- names(subset_df)
df %>% select(names_sub_df)


Error: All select() inputs must resolve to integer column positions.
The following do not:
*  as.vector(names_sub_df)

Here is a reproducible example:

key <- c("a1", "a2", "a3", "a4", "a5", "a6", "a7", "a8", "a9", "a10", "a11", "a12", "a13", "a14", "a15", "a16", "a17", "a18")

value <- c("G", "CTT", "C", "C", "G", "C", "T", "C", "C", "C", "G", "T", "C", "G", "T", "A", "T", "G")


test2 <- data.frame(key, value, stringsAsFactors = FALSE)

library(tidyr)

@Christoph please tell me why the error occurs? What am I missing? — Dr. Richard Tennen
– Dr. Richard Tennen, Commented Nov 26, 2017 at 11:50
Possible duplicate of Pass a vector of variable names to arrange() in dplyr — Elin
– Elin, Commented Nov 26, 2017 at 11:57

Phil · Accepted Answer · 2017-11-26 11:31:56Z

2

In the absence of a minimal reproducible example using mtcars as an example.

You can wrap your subset dataframe in colnames so select uses the names, not the whole dataframe, for the subsetting:

mtcars
subset_mtcars = c("hp", "drat", "wt")
subset_mtcars = mtcars[, subset_mtcars]
subset_mtcars

library("tidyverse")    
mtcars %>% 
  select(colnames(subset_mtcars))

#                      hp drat    wt
# Mazda RX4           110 3.90 2.620
# Mazda RX4 Wag       110 3.90 2.875
# Datsun 710           93 3.85 2.320
# ...

answered Nov 26, 2017 at 11:31

Phil

4,4642 gold badges26 silver badges34 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Dr. Richard Tennen Over a year ago

thanks a lot for the answer, I am sorry for not providing the example, I thought my explanation will be enough, apologizing. Please be so kind and tell me why the error occured?

Elin Over a year ago

Because you are passing a list of quoted strings and select wants unquoted names. There are a number of answers to this question if you search.

Phil Over a year ago

@elin I don't think that's quite right; I think the OP was trying to pass a data frame to select. Admittedly that's still not the correct structure 😀

Dr. Richard Tennen Over a year ago

@Elin it worked perfectly before on my pc, when I have transferred the same code to my work pc, it failed. Will check again.

Elin Over a year ago

I would also suggest always using the namespace notation when you are moving things to different computers since you may or may not have dplyr loaded. @Phil names_sub_df <- names(subset_df) is a character vector. Here's another duplicate with a select() example stackoverflow.com/questions/33284790/….

Christoph · Accepted Answer · 2017-11-26 15:43:55Z

0

From your example I am not sure whether you are looking for selecting columns or the values in a column. If you are searching for the latter, the following will do the job:

subset_df <- c("a3", "a4", "a5")
test2[test2$key %in% subset_df, ]

answered Nov 26, 2017 at 15:43

Christoph

7,2676 gold badges48 silver badges99 bronze badges

1 Comment

Dr. Richard Tennen Over a year ago

I am searching for column names in the big dataframe, I want to subset using dplyr select.

Collectives™ on Stack Overflow

Dplyr select throws a resolve to integer error, but basic select works, why?

2 Answers 2

5 Comments

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

5 Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related