I have a table in Excel. Some cells contain texts with greek letters like "MIP-1β" (the last beta has is unicode 03B2). I often copy these tables to the clipboard and read them in R with read.delim("clipboard"). In R the greek symbols are substituted by "?". However, I can copy the text correctly to Notepad++, also here to this input box in RG (as you can see above). I tried to change the fileEncoding to UCS-2LE what did not change anything. I also tried to copy the table from Excel to Notepad++, explicitly converted the encoding to UTF-8 or USC-2LE and saved it and tried to read this in R but this did not work either.
I am sure that I make some stupid mistake somewhere. I am not an expert in this encoding and locale-stuff...
I use Windows 7 and R 2.15.1 (64bit), Platform: x86_64-pc-mingw32/x64 (64-bit).
Excel is 14.0 (32bit), Notepadd++ 5.8.7 (UNICODE).
try to imported as a csv file, and use either:
fileEncoding="ISO-8859-7"
or
fileEncoding="WINDOWS-1253"
Thanks, but that's not working either. Saving as CSV in Excel seems to remove the Unicode characters already.
I created a small sample file (unicode text, tab delimited). It is correctly shown in Notepadd++ where its encoding is given as "UCS-2-LE". I can not read this file correctly in R, no matter what fileEncodings I try.
Not working:
> x = read.delim("clipboard",fileEncoding="UTF-16")
Warnmeldungen:
1: In read.table(file = file, header = header, sep = sep, quote = quote, :
ungültige Eingabe gefunden in der Eingabeverbindung 'clipboard'
2: In read.table(file = file, header = header, sep = sep, quote = quote, :
unvollständige letzte Zeile von readTableHeader gefunden in 'clipboard'
> x
[1] X.
(oder row.names mit Länge 0)
>
hi,
The sample dataset that you provided is UTF-16.
Here is the output of the "file" command for the DataUnicode.txt:
Little-endian UTF-16 Unicode text, with CRLF line terminators"
The "file" command for Windows, is included in Rtools,
http://cran.r-project.org/bin/windows/Rtools/
Also the R-command
x1 x1
Diet Cytokine
1 ST IL1β
2 ST MIP-1α
3 ST MIP-1β
4 HF MIP-3γ
5 ST MDC
6 ST MMP9
Therefore, i have two suggestions:
1. You can convert your dataset into a csv file check it with the "file" command and import it accordingly into R
or
2. please, provide a small sample of the original excel file and i will try to help
cheers, Kostas
Have you tried using the read.delim command : "stringsAsFactors=FALSE".
hi again,
Encoding (either as a function or as an argument within another function) can take only 3 values:"latin1" or "UTF-8" or "bytes". All the other encodings are presented as "unknown". This is why you are not able to read from your "clipboard". The function read.delim("clipboard") can not have as argument the "fileEncoding" because it is not a file and you can use only the argument "Encoding". i.e. read.delim("clipboard", header= T, sep="\t", as.is = T, encoding=c("latin1","UTF-8")).
On the contrary, the argument "fileEncoding" within a function read.csv("data.txt") can take as value the "UTF-16", this why you can read your files (.txt, .csv....) into R.
cheers, Kostas
Kostas, please see the screenshot above: it does not work for me, obviousely.
Encoding() is meant to "Read or set the declared encodings for a character vector" within R workspace, not an outside file.
In any case, Encoding() never worked for me.
From the screenshot i can see that you were able to import your data fine as a file. Therefore, you do not have to alter any encoding.
cheers, Kostas
A simple example of Encoding() not working......:((
a1
No, Costas, the screenshot shows that I was NOT able to read the file properly! The reading stopped when R encountered the "gamma"-character, the returned data.frame is incomplete.
Sorry, my mistake. My German is rather non existing.
Can you please try the following command:
x1
I tried this and it gives the very same error as above. The messegas say that some "invalid input" if found in the file and that the file is not read completely. The "invalid input" obviousely is the gamma character of MIP-3[gamma].
It does not give me this error (please, see the attached screenshot) on a similar to yours system.
One of my concerns, though, is the old version of R that you are using, almost 2 years old.
The next step is to check the R environment variables.
Two commands to help you on this task:
Sys.getlocale(category = "LC_ALL") # will give you all the language variables
Sys.getenv(x = NULL, unset = "", names = NA) # will give you all the environment variables
cheers, K
Thanks for your help.
I could not identify any problems with the language and environment variables.
I will now not install another R 3.0.x version, I rather wait for 3.1.0 to be released in a couple of days. Then I will try this again. I hope you are correct and my problem will be solved. However, this remains strange that is doesn't work with 2.15 although it is supposed to work (afaik).
Update: I now have installed the R version 3.1.0, and it is still not working. All the same problems as before.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R