How can I ignore null headers in a .csv file? - r

Question

How can I ignore null headers in a .csv file?
I have a csv file like this
http://190.12.101.70/~digicelc/gestion/reportes/import/liquidacion/13958642917519.csv
and my code is
data <- read.csv('1.csv',header = T, sep=";")
So R tells me
more columns than column names
And I don't want to skip the header of the file
thank you!

Answer 1

I don't see the same behavior here. R adds default column names and NA to unavailable data.
> data <- read.csv("test.csv", header = TRUE, sep = ";")
> data
col1 col2 col3 col4 X X.1
1 val1 val2 val3 val4 val5 NA
2 val1 val2 val3 val4 val5 NA
Are you using the latest version?

Answer 2

But the error message tells you exactly what the problem is. you have more columns than column names.
download.file("http://190.12.101.70/~digicelc/gestion/reportes/import/liquidacion/13958642917519.csv", destfile="1.csv")
D1 <- read.csv2("1.csv", skip=1, header=FALSE)
firstlines <- readLines("1.csv", 3)
splitthem <- strsplit(firstlines, ";")
sapply(splitthem, length)
# [1] 28 42 42
So you have 42 data columns (separated by semicolons) but 28 column names (again, separated by column names). How would R know which name you would want to go with which column? ("Computers are good at following instructions, but not at reading your mind." - Donald Knuth).
You need to edit the source file so that each column would have a name or skip the first row and then get the column names form somewhere else.
edit
yes, the idea is to take the first names and then standart variables like V1, V2, or
whatever- Otherwise is there a way to skip those?
Ok then I would just use the above with slight modification:
download.file("http://190.12.101.70/~digicelc/gestion/reportes/import/liquidacion/13958642917519.csv", destfile="1.csv")
D <- read.csv2("1.csv", skip=1, header=FALSE)
header <- strsplit(readLines("1.csv", 3), ";")[[1]]
names(D)[1:length(header)] <- header
Now you have the first 28 variables named, and the rest named V29-V42.
You can "skip" the rest of the names in various ways. If you do as suggested in another answer (Dave) , basically as
names(D) <- header
... then variables 29-42 will have NA name. This is not a usable name, and you can address these variables only by column number. Or you can do:
names(D)[29:43] <- ""
Now you can't use these names either.
> D[[""]]
NULL
I think it is useful to give them names, as many data frame operations presume names. For example, suppose you have empty names ("" as above) and try to see the first few rows of your data frame:
head(D)
# skipped most of the output, keeping only column 42:
structure(c("-1", "70", ".5", "70", "266", "70"), class = "AsIs")
1 -1
2 70
3 .5
4 70
5 266
6 70
So when using head, you will see your data frame with funny names. Or another example:
D[1:3,29:31]
.1 .2
1 C_COMPONENTE_LIQ_DESDE_CO 243 LIQUIDACION TOPE CO
2 C_COMPONENTE_LIQ_DESDE_CO 243 RESIDUAL CO
3 C_COMPONENTE_LIQ_DESDE_CO 243 RESIDUAL CO
The first component now is named "", the second one ".1", and the third one ".2". Have a look at a quote from data.frame help file below:
The column names should be non-empty, and attempts to use empty names will have
unsupported results. Duplicate column names are allowed, but you need to use check.names
= FALSE for data.frame to generate such a data frame. However, not all operations on
data frames will preserve duplicated column names: for example matrix-like subsetting
will force column names in the result to be unique.
Or suppose you add some columns to the beginning of your data frame; if you have col names then you can still address what was previously 29th column as D$V29, but with D[,29] you will get something else.
Probably there are other examples. In other words, you can have "unnamed" columns in a data frame but I don't think it is a good idea. And technically, all columns in a data frame will always have a name (it can just be "" or NA), so why not have meaningful names? (Even V29 is better than nothing.)