[Rd] [R] read.fwf and header

[Gregor Gorjanc]

Marc Schwartz wrote:
> On Mon, 2006-10-30 at 19:51 +0100, Gregor Gorjanc wrote:
>> Hi!
>>
>> I have data (also in attached file) in the following form:
>>
>> num1 num2 num3 int1 fac1 fac2 cha1 cha2 Date POSIXt
>>  1                1   f q   1900-01-01 1900-01-01 01:01:01
>>  2 1.0 1316666.5  2 a g r z            1900-01-01 01:01:01
>>  3 1.5 1188830.5  3 b h s y 1900-01-01 1900-01-01 01:01:01
>>  4 2.0 1271846.3  4 c i t x 1900-01-01 1900-01-01 01:01:01
>>  5 2.5  829737.4    d j u w 1900-01-01
>>  6 3.0 1240967.3  5 e k v v 1900-01-01 1900-01-01 01:01:01
>>  7 3.5  919684.4  6 f l w u 1900-01-01 1900-01-01 01:01:01
>>  8 4.0  968214.6  7 g m x t 1900-01-01 1900-01-01 01:01:01
>>  9 4.5 1232076.4  8 h n y s 1900-01-01 1900-01-01 01:01:01
>> 10 5.0 1141273.4  9 i o z r 1900-01-01 1900-01-01 01:01:01
>>    5.5  988481.4 10 j     q 1900-01-01 1900-01-01 01:01:01
>>
>> This is a FWF (fixed width format) file. I can not use read.table here,
>> because of missing values. I have tried with the following
>>
>>> read.fwf(file="test.txt", widths=c(3, 4, 10, 3, 2, 2, 2, 2, 11, 20),
>> header=TRUE)
>>
>> Error in read.table(file = FILE, header = header, sep = sep, as.is =
>> as.is,  :
>> 	more columns than column names
>>
>> I could use:
>>
>>> read.fwf(file="test.txt", widths=c(3, 4, 10, 3, 2, 2, 2, 2, 11, 20),
>> header=FALSE, skip=1)
>>    V1  V2        V3 V4 V5 V6 V7 V8          V9                 V10
>> 1   1  NA        NA  1    f  q     1900-01-01  1900-01-01 01:01:01
>> 2   2 1.0 1316666.5  2 a  g  r  z              1900-01-01 01:01:01
>> 3   3 1.5 1188830.5  3 b  h  s  y  1900-01-01  1900-01-01 01:01:01
>> 4   4 2.0 1271846.3  4 c  i  t  x  1900-01-01  1900-01-01 01:01:01
>> 5   5 2.5  829737.4 NA d  j  u  w  1900-01-01
>> 6   6 3.0 1240967.3  5 e  k  v  v  1900-01-01  1900-01-01 01:01:01
>> 7   7 3.5  919684.4  6 f  l  w  u  1900-01-01  1900-01-01 01:01:01
>> 8   8 4.0  968214.6  7 g  m  x  t  1900-01-01  1900-01-01 01:01:01
>> 9   9 4.5 1232076.4  8 h  n  y  s  1900-01-01  1900-01-01 01:01:01
>> 10 10 5.0 1141273.4  9 i  o  z  r  1900-01-01  1900-01-01 01:01:01
>> 11 NA 5.5  988481.4 10 j        q  1900-01-01  1900-01-01 01:01:01
>>
>> Does anyone have a clue, how to get above result with header?
>>
>> Thanks!
> 
> The attachment did not come through. Perhaps it was too large?
> 
> Not sure if this is the most efficient way, but how about this:
> 
> DF <- read.fwf("test.txt", 
>                 widths=c(3, 4, 10, 3, 2, 2, 2, 2, 11, 20),
>                 skip = 1, strip.white = TRUE,
>                 col.names = read.table("test.txt", 
>                                        nrow = 1, as.is = TRUE)[1, ])
> 

Argh, my fault as I forgot to attach it :(

> Not sure if this is the most efficient way, but how about this:
>
> DF <- read.fwf("test.txt",
>                 widths=c(3, 4, 10, 3, 2, 2, 2, 2, 11, 20),
>                 skip = 1, strip.white = TRUE,
>                 col.names = read.table("test.txt",
>                                        nrow = 1, as.is = TRUE)[1, ])
>

That is a very nice compromise! No need for [1, ], due to nrow=1.

> Of course, with the limited number of columns, you can always just set
>
> colnames(DF) <- c("num1", "num2", "num3", "int1", "fac1",
>                   "fac2", "cha1", "cha2", "Date", "POSIXt")
>

I fully agree here, but I kind of lack this directly in read.fwf. I hope
that someone from R-core is also listening to this ;)

Thank you!

Gregor
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test.txt
Url: https://stat.ethz.ch/pipermail/r-devel/attachments/20061030/88560f7c/attachment-0004.txt