[EpiData-list] EpiData Analysis: testing of version 2.0.1.104

epidata-list at lists.umanitoba.ca epidata-list at lists.umanitoba.ca
Thu Feb 1 12:23:42 CST 2007


Public testing of version 2 of EpiData Analysis can now begin.

A lot of work has been done to reshape internal parts of EpiData 
Analysis. All users are now encouraged to participate in the testing.
A system of comparing estimates, counting with results of Stata or Epi6 
is applied. Currently the system contains around 1000 different tests. 
But maybe the system is not taking care of all problems, therefore user 
tests are also important.

A complete list of changes is shown at 
http://www.epidata.dk/analysisversioninfo.htm.
Main points are:
Output is simplified. You only get what you ask for. To get more add 
options (e.g. /T) to the command
Tables Extended (sorting, percentages, weights, outbreak)
Metadata commands: missingvalue, labelvalue, rename
Outbreak analysis: "tab outcome exposure variables /oa" "epicurve 
datevar outcome group"

The test version is placed at: Http://www.epidata.dk/testing.php
Please discuss the new version on this list.
If you find definite bugs report to http://www.epidata.dk/php/mantis, 
where all development is documented.
..............................................................
Regarding the discussion last week of differences btw. open-source and 
free software.

As announced earlier a "master plan" for rewriting all EpiData Software 
to Open-Source as well as platform independence is part of the general 
development plan for EpiData. Releasing software as open-source makes no 
sense if the structure and content of the product is not sufficiently 
modular to allow for several independent development lines.

Someone might disagree with me on this, but knowing how fragile software 
development is and seing the increased requirements for documentation 
and quality assurance I am assured that a controlled situation is 
preferable at this point in time.  For almost all end users having a 
stable, quality assured software line is more important than having 
access to independent and possibly confusing development.  But there 
should be no doubt that the aim is at an appropriate time to release the 
product source-code as open-source.

Should anyone have a lot of energy or time to add to the project there 
are many issues needing assistance, e.g. answering questions on the 
epidata-list, assisting in sorting incoming mails to the help line, 
releasing mails for the epidata-list coming from un-subscribed persons, 
creating a FAQ collection such as that running for Stata at 
http://www.stata.com/support/faqs/ or joining a group running a wiki for 
EpiData, which could contain the FAQ. A start of this can be seen at 
www.epidata.org/wiki, but will not be announced before such a group has 
been established (A coordinator has been found). Notice also that for 
EpiData Entry users can develop their own CHK file commands. Modules for 
doing so (source code and examples) are found on 
http://www.epidata.dk/documentation.php
..............................................................

Kind regards
Jens Lauritsen
EpiData Association



More information about the EpiData-list mailing list