vignette stuff runs perfectly when installed from source from CRAN. The github stuff seems to be the only problem for me. But the CRAN stuff is great, good enough for me. Seems like a great product.

Error: processing vignette ‘XLConnect.Rnw’ failed with diagnostics:
‘x’ must be an atomic vector
Execution halted

I run Sweave/knitr all the time, so I know I have the necessary TeX goodies.
I’ve tried master, 0.2-4, 0.2-5

And yet I personally love this particular one, “XLConnect 0.

1-7 | Mirai Solutions” the very best. Thanks ,Shad

generally, in Excel every cell has a type (string, numeric, …). As I understand, the ID and SAMPLE_DATE columns are supposed to be numeric. However, in case in any of those (supposed-to-be) numeric columns a string would appear (even just a single cell), XLConnect would need to do a conversion to string – because string is the more generic type and data.frame columns in R can only be of one single type. A side-effect of the numeric-to-string/character conversion in the current version of XLConnect is that you will see a “.0″ appended to integer numbers. What you can try to do is to force a conversion to numeric for those columns. Any cells that are not numeric and cannot be converted to a number (because they are strings like “asdf”), would be converted to NA – but you would easily be able to detect those. Also, if you don’t mind and if it’s inline with your company’s policies, I would ask if you could send me the file (or a sample subset) to xlconnect@mirai-solutions.com such that I can have a look.

Best regards,
Martin

According to Excel all five columns within the short test file (test.xlsx) have a type of text, yet two of the columns that could be interpreted as numbers are being modified before I get them. I have tried using colTypes = ‘character’ and the data frame does indeed have all character data types (shown below), but the strings look like Excel changed them into numbers. (According to my understanding and testing forceConversion = TRUE is not doing anything, but it should not hurt.)

Note: If I insert an apostrophe at the beginning of each cell within the ID and SAMPE_DATE columns, the conversion to numbers does not occur. However, I am getting these files from another person, there are thousands of rows and I need an automated mechanism.

From the file.
ID SAMPLE SAMPLE_DATE TEST SPECIES
12345 P12345 20090408 POSITIVE MOUSE
54321 P54321 20090324 POSITIVE MOUSE
Output from R session

> test_wkbk test_df test_df
ID SAMPLE SAMPLE_DATE TEST SPECIES
1 12345.0 P12345 2.0090408E7 POSITIVE MOUSE
2 54321.0 P54321 2.0090324E7 POSITIVE MOUSE
> str(test_df)
‘data.frame’: 2 obs. of 5 variables:
$ ID : chr “12345.0″ “54321.0″
$ SAMPLE : chr “P12345″ “P54321″
$ SAMPLE_DATE: chr “2.0090408E7″ “2.0090324E7″
$ TEST : chr “POSITIVE” “POSITIVE”
$ SPECIES : chr “MOUSE” “MOUSE”

readWorksheetFromFile also requires the specification of the sheet you actually want to read in. Please see the help file for more information and/or have a look at the package vignette at http://cran.r-project.org/web/packages/XLConnect/vignettes/XLConnect.pdf

With respect to quietly = TRUE: according to the help file of require that only prevents printing of package loading confirmation messages. I guess what you are looking for is suppressPackageStartupMessages.

Hope that helps.

Best regards,
Martin

> require(XLConnect, quietly = TRUE)
Loading required package: rJava
XLConnect 0.2-3 by Mirai Solutions GmbH
http://www.mirai-solutions.com ,
http://miraisolutions.wordpress.com

> positive_animal_file positive_animals <- readWorksheetFromFile(positive_animal_file)
Error in function (classes, fdef, mtable) :
unable to find an inherited method for function ‘readWorksheet’ for signature ‘"workbook", "missing"’

Do you have any suggestions? The builds seemed to be fine.
P.S. Why does quietly = TRUE not work?