Kelly wrote yesterday that it would be relevant to create menu systems
such as in Epi6.
One way to do this would be to create in EpIData Analysis a new html file.
Look at the info.htm supplied with the test version now available at
http://www.epidata.dk/testing.php
As part of this page there are three links looking like this:
Examples of how the programme works:
Graph: "Bar chart" (Click) "Statistical Process Control Chart" (Click)
"Descriptive Means and descriptives" (Click)
If you load the info.htm file into the editor of EpiData Analysis (press
Viewer on the process toolbar and "edit HTML output" then you will see
that behind the three links are actual commands, e.g. the link behind
"Bar Chart" Contains:
<A HREF='epi:CLS;close;set echo=off;define fn __________________________
glob;fn= "@sysdir"+"testdata";cd "@fn";read bromar;set warning=on;bar
agegrp;cd; set command=off;echo Press F1 to view info again;set
echo=on'>Bar chart</a> (Click)
So even this looks a bit complicated you can see that it is ordinary
commands, such as
define fn ____________ global
read bromar
bar agegrp
So to build a menu system that does analysis you could either have in
the link:
<a href=epi:'run myprogram.pgm'>do the analysis</a> or add the actual
commands contained in the "myprogram.pgm" file separeted by semicolons.
Users are encouraged to experiment with this type of developments.
Examples could then be placed on the examples page of epidata.dk
To facilitate such menu or analysis systems EpiData Analysis has been
made such that almost all parts of the interface can be turned off, e.g.
the main menu, toolbars or the command prompt.
Jens Lauritsen
Initiator and Coordinator for EpiData
EpiData Association
Greetings. 2 things.
1. I saw an example from Epidata which looked like this:
c:\epidata\epidatastat myreport.pgm
I haven't tried this but, does this mean that epidatastat will execute
the .pgm???
I would lke to see all components of EpiData, analysis, editor etc. to
be
Able to for example: EpiENTER <file.rec>, EpiEDITOR <*.pgm>, EpiSIS
<myreport.pgm> etc.
Why? When developing systems using MENUs it seems to be better to
'hide' from the
General user the database, reports .pgm etc. Selecting EpiData and then
option 4
Exposes databases etc. to perhaps a not so sophisticated user.
2. I work for the CDC/Global AIDS Program and I think I have managed
folks to
Investigate placing EpiData, analysis etc. into what we call a software
development
Tool kit.
I have been asked if I would conduct a 'brown bag' demonstration etc. I
do not think
I have enough experience as yet to do the software justice, expecially
Analysis. Is there someone out there who would like to discuss this
with me?
Thanks and stay safe, kelly
Someone asked off line how to accomplish this
If you have to restructure files often based on the same qes file the revision of structure can be saved as EpiC commands in a batch or command file (e.g. revise.cmd)
See the documentation for EpiC for further instructions, e.g.:
EpiC revise newqes.qes datafile.rec FIRST FORCE LOG="what_happended.log"
Would revise the file "datafile.rec" based on contents of "newqes.qes" and force restructuring regardless of this would mean loosing fields (Option FORCE). The results are saved to the log file "what_happended.log"
EpiC can do other parts also, e.g. create or open a zip archive or encrypt (decrypt) files.
EpiC is available from the download page of http://www.EpiData.dk
Example: Assume you recieve files with administrative data every month called datafile1.txt datafile2.txt datafile3.txt
To import these data, make control lists and prepare a reports you do this:
Create
- a file called structure.qes which contains the same format for fields as the txt file with administrative data.
- a file called newstructure.qes containing the fields you wish to keep in an approriate format.
- a file called report.pgm which will in analysis append the three files and make a standard report.
Then create the file revise.cmd with this content:
*.......revise.cmd ....................................................................
EpiC revise newqes.qes datafile.rec FIRST FORCE LOG="what_happended.log"
EpiC import TXT datafile1.txt qes=structure.qes out=datafile1.rec delim=; q=text ignorefirst date=yymmdd
EpiC import TXT datafile2.txt qes=structure.qes out=datafile1.rec delim=; q=text ignorefirst date=yymmdd
EpiC import TXT datafile3.txt qes=structure.qes out=datafile1.rec delim=; q=text ignorefirst date=yymmdd
EpiC revise newstructure.qes datafile1.rec FIRST FORCE LOG="what_happended.log"
EpiC revise newstructure.qes datafile2.rec FIRST FORCE LOG="what_happended.log"
EpiC revise newstructure.qes datafile3.rec FIRST FORCE LOG="what_happended.log"
* then you could make a list of data from the centers involved:
EpiC count center datafile1 datafile2 datafile3 LOG="centerlist"]
*You could also export datafile1 to a spreadsheet format for the center with number 276 for particular checks:
EpiC e xls datafile1.rec datacenter276.xls filter="CENTER=276"
* and finally you could make a standard report using analysis:
c:\program\epidata\epidatastat myreport.pgm
.............................revise.cmd to here
You would then only have to start the revise.cmd to create your monthly report.
Please note that now all previous postings to EpiData are available in searchable form from:
http://lists.umanitoba.ca/pipermail/epidata-list/
Jens Lauritsen
Initiator and Coordinator of EpiData
EpiData Association
Someone asked off line how to accomplish this
Situation:
You have a file with say 1000 observations and you wish to encrypt the field containing social service numbers:
1. use epidata to make a qes file (tools menu - qes file from recfile)
2. edit the field in question from
e.g.
string field : v1 social id __________________________
to v1 social id <e > (same length)
numerical: v1 social id ########################
to v1 social id <e >
3. Save the qes file
4. start enter data (point 4 on work process toolbar) and EpiData asks if you wish to update based on the revised qes file.
5. Do so and you are asked for passwords
Thats all.
Jens Lauritsen
Initiator and Coordinator of EpiData
EpiData Association
>>> <epidata-list(a)lists.umanitoba.ca> 31-03-2005 16:52:43 >>>
>I use EPIDATA software occasionally and everything works fine in Windows98 and Windows 2000.
> But in Windows XP or XP-Pro when run as normal User (NOT Administrator) , EPIDATA refuses to open files (QES or REC) While running as ADMINISTRATOR, everything works OK !
Setting up user rights in Windows XP can be very troublesome. It is not possible to give general advice for this, since there are very many local policies. In some places users have no rights at all, in other restricted access to one folder, ......
The EpiData programme as such does not read any of these rights and can be freely copied from one folder, memory stick or CD-rom or moved without "real installation", since the registry access is not part of the installation (unless you decide to associate file types with epidata).
To work with rec and qes files the users need write access rights in that folder. So instruct the users to work on their default data folder - e.g. this could be "C:\documents and settings\mdi256\documents". But to share files btw. users you would need common rights to that folder. Also the file "epidata.ini" is saved after running epidata with current settings of screen, options etc. So the user needs write rigths for that or an error message box will appear at closing of EpiData.
So in general:
a. Let each user install epidata in their own data space (it is only around 2-4Mb for the whole programme) and each user can work independently.
b Set up your pc such that each user have sufficient rights
Jens Lauritsen
Initiator and Coordinator of EpiData
EpiData Association
> I have a questionnaire entered in Epi Data 3.0 and want to analyze it in
> Epi Info 3.2.2. When I bring in my data I lose all the value-lables
> that I had defined using the command "label-block" in Epi Data. I want
> to know what to do to keep these labels.
> Sylvie Delcroix, France
Epi Info for windows from CDC does not use labels. This is part of the reason I started EpiData, because in analysis the labels are crucial as documentation.
If you want to use the labels in analysis you will have to use EpiData Analysis which is getting closer to release. If your analysis is general crosstables, descriptive statistics and graphs the test version available at http://www.epidata.dk/testing.php will do the work for you.
You could decide the strategy:
1. Do the basic analysis you have to do anyway with EpiData Analysis.
2. If you need to do further advanced analysis do the rest in Epi Info for windows (or other software)
Regards
Jens Lauritsen
EpiData Association