Ftyjtk

such data frame as follows:

df <- data.frame(yr = rep(2000:2017,each=46),

                 dayorder = rep(seq(1,365,8),time=18),

                 fid = rep(1:46,time=18),

                 Value = runif(46*18)

                 )

when I sieve by keys:

df[which(dayorder==9),]

      yr dayorder fid      Value

2   2000        9   2 0.3424053

48  2001        9   2 0.6639720

94  2002        9   2 0.5530076

140 2003        9   2 0.9757845

.....

When use 'yr' key ,report error:

df[which(yr==2001),]

Error in which(yr == 2001) : object 'yr' not found

same error by 'fid' key"

df[which(fid==2),]

Error in which(fid == 2) : object 'fid' not found

but df$yr is existed:

df$yr

  [1] 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000

 .....

What's the problem of data frame?

You need to add df$ before the column name, in the first case you don't get an error probably because dayorder is a variable defined in your environment.

It should've thrown an error like the others.

df[which(df$dayorder==9),] # or df[df$dayorder==9,] 

When I call like this df[which(df$dayorder==9),] I get:

Error in which(dayorder == 9) : object 'dayorder' not found

Another solution using with, avoids too many df$ (also note you can remove which():

with(df, df[dayorder==9, ])

you can try a tidyverse as well

library(tidyverse)

filter(df, yr == 2001)