dhis2-devs team mailing list archive

[Jason Pickering <jason.p.pickering@xxxxxxxxx>]

Hi Craig,

Its hard to say, but it sounds like a lot. The system could probably cope
however, but I am not sure I would design the system this way.

Not knowing anything about the data, is the project code something very
specific? I am thinking that a given project code would have a specific
team and implementing partner.  Is it so that a given project code can have
multiple teams and implementing partners? If the team and partner can be
derived from the project code (and this is managed externally) then you
would only need people to select the project code, then the partner and
team could be "derived". This might also help to improve accuracy of the
data, as it might be very easy for people to select an invalid option when
doing data entry.  If there is a multi-to-multi sort of relationship
between all of this, then I think you could design it similar to the "ART
monthly summary" in the demo database (with one more dimension for team),
but it would be a lot simpler if you could derive partner and team from the
project code.

If you could do it this way (with only project code) , this is quite
similar to another very large system which is using DHIS2, and has
thousands of "project codes" or mechanisms.  These can also can be further
restricted to users using sharing, assigning an organisation unit to the
category option, and specifying a valid start and end date, to decrease the
amount of available options to users.

As for dis-aggregating by target/results, this is really not something you
should do with the data set category combinations unless you are really
careful I think.  The reason is that by default in the analytics, targets
will be added to results, and that is generally not what you want. Likely,
you would need to create two classes of data elements: one for targets and
another for results. You could use the cat combos to do this, but all users
would need to understand that the default action in the analytics would be
to add everything together.



Hope this helps.

Best regards,
Jason


On Thu, Oct 22, 2015 at 12:40 AM, Craig Hollingsworth <
craig.hollingsworth@xxxxxx> wrote:

> Hi,
>
> Good work on the new release.  Looks cool!
>
>
> I'm exploring different config options for a possible system for NRC.
> I'm looking at datasets with 4 attributes (project code, team, implementing
> partner and target/actual.
>
> The potential number for each option is (200 new per year, 9, 10, 2).
>  That's a spectacular amount of combinations - especially after a few years
> of use.
>
> Can Dhis2 cope?!!!!   Any advice would be greatly appreciated.
>
> I recognise that I could cut some options back and go for a slightly
> different config, but the above would be the preferred/easiest option for
> us.
>
>
> Cheers
>
>
> Craig
>
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~dhis2-devs
> Post to     : dhis2-devs@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~dhis2-devs
> More help   : https://help.launchpad.net/ListHelp
>
>


-- 
Jason P. Pickering
email: jason.p.pickering@xxxxxxxxx
tel:+46764147049