Parse complex text file in R

Question

I am looking to parse a text file in R to be loaded as a data.frame. I have a long text file with fixed width data, seperated by sections (ID) and subsections (SUB). The length of each section is variable. I'm looking to create two data frames, one for the ID section and one for the SUB section. The example data is as follows:

Header 1
METRIC       0.30    10.00
ID      K0107050 Aa
    0.06    15.24    14.40    14.40     7.13     0.13     0.19  1
    0.17    14.35    13.57    13.57     6.40     0.12     0.18  1

SUB
    1.000   1.000  0.093  0.11  0.11 301
    1.000   1.000  0.093  0.11  0.11  61
ID      K0129050 Aa
    0.06    26.35    24.90    24.90    10.88     0.62     0.88  1
    0.15    25.35    23.96    23.96    10.93     0.55     0.74  1

SUB
    3.000   3.000  0.506  0.53  0.53 102
    4.000   4.000  0.514  0.55  0.55 118

The dataframe(s) I would like are:

DF1

Header 1    K0107050    Aa    0.06    15.24    14.40    14.40     7.13     0.13     0.19  1
Header 1    K0107050    Aa    0.17    14.35    13.57    13.57     6.40     0.12     0.18  1
Header 1    K0129050    Aa    0.06    26.35    24.90    24.90    10.88     0.62     0.88  1
Header 1    K0129050    Aa    0.15    25.35    23.96    23.96    10.93     0.55     0.74  1

DF2

Header 1    K0107050    Aa  1.000   1.000  0.093  0.11  0.11 301
Header 1    K0107050    Aa  1.000   1.000  0.093  0.11  0.11  61
Header 1    K0129050    Aa  3.000   3.000  0.506  0.53  0.53 102
Header 1    K0129050    Aa  4.000   4.000  0.514  0.55  0.55 118

I've gotten so far as to use the readLines() but get stuck after that, given the different sections in the text file. Thank you

There are many more rows after ID and SUB, and not a consistent number of rows. I've only provided the above as an example. — user2325155, May 31 '18 at 21:47
What do you want to do with the blank columns at the end for the `SUB` sections? Just fill them with `NA`? — thelatemail, May 31 '18 at 22:10

score 2 · Accepted Answer · answered May 31 '18 at 22:49

Here is the start (sorry time to bed...):

x <- readLines("myFile.txt")

library(dplyr)

bind_rows(
  lapply(split(x, cumsum(grepl("Header|Metric|ID|SUB", x))), function(i){
    i1 <- i[ i != "" ]  
    nums <- unlist(strsplit(tail(i1, -1), " "))
    res <- cbind.data.frame(Grp = i1[1],
                            matrix(na.omit(as.numeric(nums)),
                                   nrow = length(i1) - 1, byrow = TRUE),
                            stringsAsFactors = FALSE)

    res
  })
)

#                   Grp    1     2      3     4     5      6    7  8
# 1            Header 1 0.30 10.00     NA    NA    NA     NA   NA NA
# 2 ID      K0107050 Aa 0.06 15.24 14.400 14.40  7.13   0.13 0.19  1
# 3 ID      K0107050 Aa 0.17 14.35 13.570 13.57  6.40   0.12 0.18  1
# 4                 SUB 1.00  1.00  0.093  0.11  0.11 301.00   NA NA
# 5                 SUB 1.00  1.00  0.093  0.11  0.11  61.00   NA NA
# 6 ID      K0129050 Aa 0.06 26.35 24.900 24.90 10.88   0.62 0.88  1
# 7 ID      K0129050 Aa 0.15 25.35 23.960 23.96 10.93   0.55 0.74  1
# 8                 SUB 3.00  3.00  0.506  0.53  0.53 102.00   NA NA
# 9                 SUB 4.00  4.00  0.514  0.55  0.55 118.00   NA NA

My apologies, I need two dataframes, one ID and one SUB. I have edited my post accordingly. — user2325155, May 31 '18 at 22:58
@user2325155 Sorry, as I said this is just a start, parsing irregular files is not fun, good luck. — zx8754, Jun 01 '18 at 06:23

Parse complex text file in R

1 Answers1