I don't know how Rbase works with Excel, but I can list some information from the other side of the fence.
Excel files are very hard to work with to import data, being that data types can range cell by cell. Using any language that supports COM objects, and Oterro, you can import data from excel sheets, but the import has to be semi hard-coded due to the aforementioned reason. Microsoft's driver for Excel takes the first 7 rows of each column and guesses the datatype. If the data in a column varies at all (ie 12/20/01 and Dec 20 2001, etc), you will encounter problems. One thing that always trips up excel importation is column headers. Being that they are actually data cells, get rid of them, it will foul up any import with garbage data/data types. > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On > Behalf Of Texmaster Express > Sent: Monday, October 08, 2001 1:39 PM > To: RBase List Server > Subject: The Data Conversion from Hell > > Good afternoon. > > I'm still working on the data conversion routine from Hades. I've decided > to > use the Gateway command since it seemed to work once before. I've saved my > Excel file to an XLS3 format (am using Excel 2002) and run the command: > > gateway import xls3 ords.xls replace orders > > I've also tried a variant of the above, > > gateway import xls3 ords.xls create x > > In each instance, R:Base returns the error message: > > Cannot import a table with 0 columns > > I've checked my source table in Excel and it reads just fine. I've also > checked to be sure I have an equal number of columns. > > I've also tried saving the file as a dbf, attaching it and projecting it, > but the Excel file has 169 columns, and the dbf file will only allow 128. > > Any suggestions? (Besides convincing our software company to use Oterro, > which I've already tried) > > Dan
