Unused arguments error on .local() when loading from file.

[Fabio Eitelvein]

Hi there,

I am trying to open a big file (~3GB) using sqlfd. I've just installed sqldf 0.4.9 on R version 3.1.1 for Mac OS X 10.9.2.

While trying to open the file like this:

f <- file("preprocessed.csv")
df <-sqldf("select * from f")

I get the following error:

Error in .local(conn, name, value, ...) :
unused arguments (comment.char = "", quote = "\"")

The same happens when I use read.csv.sql().

What am I doing wrong?

Thank you.

Best regards

[Gabor Grothendieck]

On Thursday, October 30, 2014 5:10:11 PM UTC-4, Fabio Eitelvein wrote:

Hi there,

I am trying to open a big file (~3GB) using sqlfd. I've just installed sqldf 0.4.9 on R version 3.1.1 for Mac OS X 10.9.2.

While trying to open the file like this:

f <- file("preprocessed.csv")
df <-sqldf("select * from f")

I get the following error:

Error in .local(conn, name, value, ...) :
  unused arguments (comment.char = "", quote = "\"")

The same happens when I use read.csv.sql().

Those arguments are not available.  sqldf and read.csv.sql uses sqlite to read in the data and sqlite is more limited than read.table in the range of arguments available.  

See ?sqldf

[Gabor Grothendieck]

In looking at this further there is an additional difference in RSQLite 1.0.0 (relative to the prior version) that was recently uploaded to CRAN that causes this.  

A fix is in googlecode repo at http://sqldf.googlecode.com .

Until this fixed version is on CRAN you can:

   1. build the new version from source, or

   2. source the fixed sqldf.R on the fly like this:

   library(sqldf)

   source("http://sqldf.googlecode.com/svn/trunk/R/sqldf.R")

   ... your code ...

[Fabio Eitelvein]

Hi Gabor,

On Friday, October 31, 2014 9:25:02 AM UTC-2, Gabor Grothendieck wrote:

In looking at this further there is an additional difference in RSQLite 1.0.0 (relative to the prior version) that was recently uploaded to CRAN that causes this.  

A fix is in googlecode repo at http://sqldf.googlecode.com .

Until this fixed version is on CRAN you can:

   1. build the new version from source, or

   2. source the fixed sqldf.R on the fly like this:

   library(sqldf)

   source("http://sqldf.googlecode.com/svn/trunk/R/sqldf.R")

   ... your code ...

This workaround did the trick for me.

Thank you very much! 

[fran...@gmail.com]

>    library(sqldf)
>    source("http://sqldf.googlecode.com/svn/trunk/R/sqldf.R")
>    ... your code ...

Great workaround, thanks.

[Gabor Grothendieck]

Note that a new version of sqldf was uploaded to CRAN in November so provided you have up to date R and packages then this should no longer be needed.  See Latest News on sqldf home page for more info.  http://sqldf.googlecode.com

 

[jklk...@gmail.com]

Gabor Grothendieck於 2014年10月31日星期五 UTC+8下午7時25分02秒寫道：

I have same problem but it doesn't work for me.
What should I do?

[Gabor Grothendieck]

You will need to post a reproducible example.  It works for me.

> cat("a,b\n1,2\n", file = "processing.csv")

> library(sqldf)

> f <- file("processing.csv")

> sqldf("select * from f")

  a b

1 1 2

> read.csv.sql("processing.csv")

  a b

1 1 2

> R.version.string

[1] "R version 3.3.0 Patched (2016-05-03 r70575)"

> packageVersion("sqldf")

[1] ‘0.4.10’

> win.version()

[1] "Windows 8.1 x64 (build 9600)"