The easiest way to do this is in Analysis.
First, in Excel: At the top of each column of data, you need to have a valid EpiData field name. Format each Excel column to match the data:
Numeric (with number of decimals) Text Date (format doesn't matter, but make them all the same)
Select and copy column headings (field names) and all data.
Then in Analysis,
read /cb define idnum ##### idnum = recnumber // if you want IDNUM=record number savedata xyz
If your data are clean, your data are now in EpiData.
Caution: if you use the test version of Analysis (the one that can read .epx files), you will also se xyz.chk. DELETE IT! It has errors. No problem with the current working version of Analysis.
With EpiData, create a QES from your REC file. Edit the QES to match the layout you want (e.g. may IDNUM comes first instead of last). Save the QES and open the REC for input. Say yes to reformat the file.
There you go. I find this easier that trying to export from Excel as CSV file and importing.
Jamie
On 2011-08-12, Bijay wrote:
Jens & Jamie,
Thank you for suggestions time to time on Epidata programming that I am working on. We are replacing existing disease database (specific disease) that is in Access plateform with Epidata. The access database was not stable over the years, so lost a chunk of data few times. Hope Epidata is the solution. I completed most programming including several calculated fields, some are very complex assiging risk factors. The data is in MS Excel. It has ~ 2000 data with ~ 100 variables. All the data fields in both (Epiadta QES and Excel) systems are identical except IDNUM is extra in Epidata.
I am looking the best way to migrate excel data into the blank epidata. May be I should convert excel to .REC file first, then import, or may be .TXT is better. I have tested with small dataset by cerating .REC file first then import. It did work.
Any suggestion would be appriciated.
Thanks
Bijay