Continued discussion on this topic:
The suggested solution I gave could be clarified:
1. Dates before year approx. 1700-1800 give problems in Analysis now, but obviously should not. Although as everyone probably know there is a shift around 15xx (I do not remember the exact year).
2. The suggestion: "SUGGESTION: Would it not be possible for Analysis to assign a standard valid value (e.g. 01/01/1900) to any invalid dates it comes across during the read process. "
Will be used in the next build of analysis, since it is very easy to implement. Obviously with a message to the user. I think this can be done in 5 minutes of work time in the programming, whereas the optimal solution of showing variable/field name and record is a large job - could be 1-2 days of work, since the function controlling the date value is not knowing neither the name of the field nor the record. But I will add the issue to the work plan for a solution at some point.
Regarding the "excluded" versus "deleted" is also easy. The user can control whether records deleted are read or not by: set read deleted=on set read deleted=off
The change works only when reading the file, so if you change re-read the file again. e.g. set read deleted = off read myfile describe age set read deleted = on read myfile /close describe age // here records marked for deletion are included.
off is default. There is an error in v1 b68 in relation to this setting, such that if you read the file with "set read deleted=off" and then do an "update", you will see the records marked for deletion. The error has been found and a solution will be shown in next test release.
regards Jens Lauritsen