[EpiData-list] Autobackup-feature and current test of EpiData Analysis b160

epidata-list at lists.umanitoba.ca epidata-list at lists.umanitoba.ca
Mon Feb 23 16:22:01 CST 2009

The development is now at a crucial point in time getting to a certain 
"gate". Before the gate is opened EpiData Analysis will be released as 
v2.2 and some updates to EpiData Entry will also take place, e.g. 
hopefully fixing the reported "auto backup" problem reported.

After the gate there will be a common core system for reading, writing 
and interpreting files for as well EpiData Entry, EpiData Analysis and 
EpiC. This will not be seen immediately by the end user, but will assure 
that all functions are doing the same regardless of which part of 
EpiData Software and errors only needs one place for bug fixes. The 
translation process will also be adapted and released in common for the 
three software parts, since after that all functions and commands in 
relation to reading and writing data files and chk files will have the 
same number code. The list and translators will be informed when 
possible on all of this later this quarter or beginning of next quarter.

Meanwhile EpiData Analysis b161 has been placed for testing for download 
from the testing.php page of the main site. (by a mistake the first were 
called b160, but actually were b161) - if you report problems report the 
date of the build if this was b160).

One new aspect is that one commented by J. Hockin in the Mantis issue 
number 335. (A template for SPC graphs in a menu like system): The 
technical principle is the following:

A: An htm file is created by the user outside of analysis with a 
combination of EpiData Analysis commands and ordinary html coding. (here 
called spc.htm)
B: This file is activated in the help window via an entry in the 
graph-spc toolbar menu/dialog.
       (the same as the command: helpview "spc.htm")
C: When the user clicks on a given link in the file, certain commands 
are issued and information can be given in the output window - without 
closing the helpview file.

J.Hockin argues in the Mantis issue that the strange window is difficult 
to close again and that it would be preferred to have the spc.htm file 
opened in the main results window. I experimented a bit with both 
solutions and it turns (to me) an easier approach to use the help window.

This is the same approach used if one uses the menu "Windows->screen 
font". Users can develop other htm files for such menu systems and load 
them either into help window (helpview spc.htm) or result window (view 
spc.htm) and if you wish an easy return to such a file issue the "set 
START PAGE " command after which you can open your menu file with F8 or 
have the system open it at start of the exe file. For inspiration on 
creation of such files, see the startfont.htm and the spc.htm files.

Apart from testing that new menu/command button principle it is 
important for us to get tests of the Statistical Process Control Graphs 
and the new features introduced. I will contact a number of users 
directly, but all are more than welcome to participate.

The optimal is if users would :
a1: Create simple datasets using the built in system accessible via the 
toolbar-graph-spc menu/dialog in such a way that the following can be 
- freeze (where estimates are based on the first n data points
- breaks (where the time series are split in sub-parts)
- exclude (where individual data points are excluded from estimates)
and the tests (t1-t5), which will show different tendencies/aberrations.

The optimal is to send such datasets in a zip file to "info at 
epidata.dk" consisting of the data and the pgm file to run the tests, 
including expected results. Disregard the mail notion of the mail not 
being received - we will let these through the spam filter and contact 
you directly.

regards Jens Lauritsen
EpiData Association

More information about the EpiData-list mailing list