From today v0.9.3.0 is released at the bottom of :
Http://www.epidata.dk/testing.php
Notice that the commands.html and the "how to adapt your pgm" help
files found in the help menu of the main window.
Please test and report to this list or in case of severe problems send
an email to info at epidata.dk. We do hope that there are no more bugs
in the released test version. But we do also encourage you to report
such errors here.
More info below
Best wishes
Jens Lauritsen , Torsten Bonde Christiansen
EpiData Association
Denmark
-------------------------------------------------------------------------------------------------------------------
Example: read a file change value in a variable and display a browser
before and after:
// ----example:
read "myfile.epx";
new variable v12 s;
v12:= "unknown";
browse age v12 !caption:="Data before change";
v12 := iif( age[_n] > 18, "adult", "young");
browse age v12 !caption:=("Data after change" + now());
// ... end
Here it would be natural to ask, why do I need the [-n]. This is a
consequence of the way the internal machinery (parser) works. We will
think over and attempt to see whether it is possible to leave out this,
but until further notice you need that part. We have made it easy to add
the [_n] - if you press alt+n with focus in the command prompt or in the
editor the sequence "[_n]" will be added.
Please do remember to work on copies of data. Real data should only be
used in production release versions for actual data-analysis or data
management
Changes from last experimental version are:
Documentation
* fixed errors and inconsistencies in commands.html
* fixed calling error - which would not open the commands.html in IE browser
Adapted syntax
* valuelabels: edit vl name <type> ( value, text) ( value, text) etc.
Added functionality
* automatic variables and memo fields may be created
* list project
* funtion user(), returning defined users in a project with extended access
* Altered shortcuts on MAC (Editor)
Bug fixes:
* merge had several issues - and should now be complete, including
documentation - could overwrite mergevar
* editing of !related parameters !childobs and !afterobs
* errors in certain variable expansion aspects and special combinations,
e.g. save a project without variables
* report validate, report cby , check relate could fail.
* check relate -> would indicate incorrectly that a 1-many relation was
not correct.
* createtime(...) will return sysmis with illegal time value