Greetings SK,
[This may be a repeat -- first posting didn't seem to get through]
I like to use a little program by Christian Maas to generate data, Random file generator from: http://www.chmaas.handshake.de/delphi/freeware/freeware.htm
The command line version can be run from a BAT file containing a statement such as: RndFileC.exe /R=50000 /Z=YYYY/MM/DD /N=4 /F=1,@ /C=15 /F=1,, /B /C=1 /S=14 /B /C=1 /S=14 /F=1,@ /Y=1945,2003 data.txt
For example, this generates 50,000 records with the given date format and range, plus other fields. Then import into EpiData.
I prefer to set up the EpiData files, enter a few records, and then export them to a CSV file; then make sure you use the same format to generate the random records to import back into EpiData.
I have use Maas's program to generate 8,000,000 records (but not this number for EpiData; this was data to analyze using gawk in GNU/Linux). The sort program is real nice, too.
Hope this helps,
Pete Geddes geddes@ix.netcom.com
----- Original Message ---------------
Subject: [EpiData-list] Program simulated data in EpiData (Analysis) From: epidata-list@lists.umanitoba.ca Date: Sun, 25 May 2008 09:38:58 -0400 To: epidata-list@lists.umanitoba.ca
Dear All I was wondering if EpiData Analysis could do simulations: here is what I am contemplating doing; I have created a database of admissions/discharges from a hospital using a standardized register and using EpiData as the software for data entry and ana;ysis. I wish to populate the fields (15 in total, of various types: boolean, categorical with k values, numeric, and date fields) I have created with say N=5,000 observations. The reason I wish to do this is to show the stakeholders (in this case Hospital Managers and Unit Heads) that one could use simple yet effective data entry and analysis software (like EpiData) to monitor and evaluate hospital-based programs. I wanted them to see EpiData in action with the simulated data--starting with EpiData Entry and then with Analysis.
Can simulation be done in Analysis? or should I prepare the dataset with Stata or SAS and then import in EpiData--I am not sure how to do this, but I will search around to see how to get it done?
Thanks, SK