Encoding in R: How to convert this string to UTF-8?

Question

I am using R to read data from an old fame database. This works fine in general but I get unexpected encoding back when reading descriptions. E.g.:

a <- "\U3e34653c"
# is supposed to be 
"ä"

I tried to iconv my self around this problem but despite trying numerous possibilities I was not able to get it displayed in a proper way. my locale: en_US.UTF-8. Is there a way around replacing (sub) such strings?

I had a similar issue and had success converting first to 'latin1'. Might be worth giving it a shot — Ricardo Saporta, Mar 18 '13 at 18:27
This seems to be caused by your database, not R. The character ä is 0x000000e4 in Unicode, not 0x3e34653c (which is undefined). — Nick Ulle, Nov 14 '14 at 05:55

Simon O'Hanlon · Answer 1 · 2013-03-19T18:11:11.877

0

Try opening the files with a different encoding string? As Ricardo suggests, perhaps Latin1? If not maybe some other exotic flavours:

f <- file( "myfile.db" , encoding = "Latin-1" )
dat <- readLines( f )

Can you link to some data?

edited Mar 19 '13 at 18:11

answered Mar 18 '13 at 17:06

Simon O'Hanlon

58,647
14
142
184

I know it ain't the right unicode character. The problem is that I just get the character (whatever it is?) I have posted when I read from the db. Currently I am trying to figure out what I could do instead of regexp replacing. – Matt Bannert Mar 19 '13 at 10:37

Alex Ioannides · Answer 2 · 2014-10-16T13:10:07.983

-1

I had an identical problem when extracting data from SQL Server (via ODBC and the RODBC package). I solved it by changing the settings on the ODBC driver to treat all strings as unicode.

More specifically, I'm using Actual Technologies ODBC driver for SQL Server and under 'Advanced Language Settings' can specify 'Treat text types as Unicode' with an option for 'Multi-byte text encoding' to be set to UTF-8.

edited Oct 16 '14 at 13:10

answered Oct 16 '14 at 12:42

Alex Ioannides

1,204
9
10

Encoding in R: How to convert this string to UTF-8?

2 Answers2

Linked