R - merge a list of data frames into one data frame with missing values by row

问题

I have a variation on the oh-so-common problem of how to merge things together in R.

I have a set of .txt files in a particular folder, and I have written a function that:

makes a list of the files I want, and then for each file
reads the file
subsets the data (to extract just the rows and columns of interest)
does some calculations on the data
adds these new values to a list.

What I end up with is a list with the following structure:

>str(DataList)
List of 16
 $ :'data.frame':   14 obs. of  2 variables:
  ..$ Sample: Factor w/ 14 levels "Sample_1A","Sample_1B",..: 1 2 3 4 5 6 7 8 9 10 ...
  ..$ Var1  : num [1:14] 27.9 33.8 29.9 29.4 28.8 ...
 $ :'data.frame':   14 obs. of  2 variables:
  ..$ Sample: Factor w/ 14 levels "Sample_1A","Sample_1B",..: 1 2 3 4 5 6 7 8 9 10 ...
  ..$ Var2  : num [1:14] 24.6 27 26.8 26.7 27.2 ...
 $ :'data.frame':   12 obs. of  2 variables:
  ..$ Sample: Factor w/ 14 levels "Sample_1A","Sample_1B",..: 1 2 3 4 5 6 7 9 11 12 ...
  ..$ Var3  : num [1:12] 31.4 35.6 34 35.7 32.5 ...

For each variable (Var1, Var2, Var3, ...) I have a column Sample and a column of numerical values.

Sample is always a factor with 14 levels; these levels are the same for each variable.

The problem is that some variables (like Var3 above) don't have observations for each level of Sample.

What I want to end up with is a data frame with 14 rows (one for each level of Sample). The first column should be Sample; then for each variable, there should be a column containing the corresponding numerical values, like so:

Sample     Var1    Var2    Var3
Sample_1A  27.9    24.6    31.4
Sample_1B  33.8    27      35.6
...
Sample_3B  26.8    29.7    NA

I've been trying to do this with do.call, but don't know how to pass the arguments for by; cbind gets unhappy because of the missing values. Any thoughts on how to do this?

Thanks!

EDIT: As per joran's request:

>dput(DataList[1:3])
list(structure(list(Sample = structure(1:14, .Label = c("Sample_1B", "Sample_1C", "Sample_1D", "Sample_2C", "Sample_2A", "Sample_2D", "Sample_3B", "Sample_3C", "Sample_3A", "Sample_3D", "Sample_4B", "Sample_4C", "Sample_4A", "Sample_4D"), class = "factor"), Var1 = c(26.9333333333333, 29.17, 28.9366666666667, 28.9233333333333,  28.61, 28.63, 26.7933333333333, 34.6633333333333, 30.4966666666667, 28.4433333333333, 27.4533333333333, 28.3, 27.9633333333333, 27.2366666666667)), .Names = c("Sample", "Var1"), row.names = c(NA, -14L), class = "data.frame"), structure(list(Sample = structure(1:14, .Label = c("Sample_1B",  "Sample_1C", "Sample_1D", "Sample_2C", "Sample_2A", "Sample_2D", "Sample_3B", "Sample_3C", "Sample_3A", "Sample_3D", "Sample_4B", "Sample_4C", "Sample_4A", "Sample_4D"), class = "factor"),                                       Var2 = c(24.19, 26.6033333333333, 26.0366666666667, 27.6766666666667, 27.61, 27.5633333333333, 25.1566666666667, 33.7266666666667, 27.7, 26.1466666666667, 25.65, 26.3633333333333, 25.5333333333333, 26.1733333333333)), .Names = c("Sample", "Var2"), row.names = c(NA,  -14L), class = "data.frame"), structure(list(Sample = structure(c(1L, 2L, 3L, 4L, 5L, 6L, 7L, 9L, 11L, 12L, 13L, 14L), .Label = c("Sample_1B", "Sample_1C", "Sample_1D", "Sample_2C", "Sample_2A", "Sample_2D", "Sample_3B", "Sample_3C", "Sample_3A", "Sample_3D", "Sample_4B", "Sample_4C", "Sample_4A", "Sample_4D"), class = "factor"), Var3 = c(31.4133333333333, 35.56, 33.9666666666667, 35.66, 32.4633333333333, 31.99, 31.3133333333333, 36.34, 34.9433333333333, 34.5433333333333, 34.3766666666667, 33.28)), .Names = c("Sample",  "Var3"), row.names = c(NA, -12L), class = "data.frame"))

回答1:

Looks like a textbook use case for Reduce.

merge.all <- function(x, y) {
    merge(x, y, all=TRUE, by="Sample")
}

output <- Reduce(merge.all, DataList)

来源：https://stackoverflow.com/questions/17171426/r-merge-a-list-of-data-frames-into-one-data-frame-with-missing-values-by-row

标签

merge

dataframe

do.call