It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.
I think it will work quite well to just use the same data element in different datasets with different periodicity. The only thing to look out for is to avoid double counting, e.g. watch out for misleading Totals in Analytics.
It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.
It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.
Hi Knut
One way to avoid double counting is to make sure both data sets are never active in the same org units.
JM
El 14/07/2016, a las 2:10 p.m., Knut Staring knutst@gmail.com escribió:
Hi David,
I think it will work quite well to just use the same data element in different datasets with different periodicity. The only thing to look out for is to avoid double counting, e.g. watch out for misleading Totals in Analytics.
It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.
22.2.5. Data elements assigned to data sets with different period types
Data elements should not be assigned to two separate data sets whose period types differ. The recommended approach would be to create two separate data elements (for instance a monthly and yearly data element) and assign these to respective datasets.
I am sure Lars can provide the exact reason for this, but I am pretty sure one should not have the same data element with two separate frequencies. It would seem to be a good feature however, but I believe this is a restriction imposed by analytics at the moment.
Regards,
Jason
···
On Thu, Jul 14, 2016 at 4:28 PM, Knut Staring knutst@gmail.com wrote:
Thanks JM, I forgot to mention that.
Though you may still have to watch the totals if you combine months and years in the same Pivot Table.
Hi Knut
One way to avoid double counting is to make sure both data sets are never active in the same org units.
JM
El 14/07/2016, a las 2:10 p.m., Knut Staring knutst@gmail.com escribió:
Hi David,
I think it will work quite well to just use the same data element in different datasets with different periodicity. The only thing to look out for is to avoid double counting, e.g. watch out for misleading Totals in Analytics.
It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.
Hm, my instinct was to say the same as the manual says, but when I tested it with the same restrictions that Juan Manuel pointed out, it seemed to work fine.
22.2.5. Data elements assigned to data sets with different period types
Data elements should not be assigned to two separate data sets whose period types differ. The recommended approach would be to create two separate data elements (for instance a monthly and yearly data element) and assign these to respective datasets.
I am sure Lars can provide the exact reason for this, but I am pretty sure one should not have the same data element with two separate frequencies. It would seem to be a good feature however, but I believe this is a restriction imposed by analytics at the moment.
Regards,
Jason
On Thu, Jul 14, 2016 at 4:28 PM, Knut Staring knutst@gmail.com wrote:
Thanks JM, I forgot to mention that.
Though you may still have to watch the totals if you combine months and years in the same Pivot Table.
Hi Knut
One way to avoid double counting is to make sure both data sets are never active in the same org units.
JM
El 14/07/2016, a las 2:10 p.m., Knut Staring knutst@gmail.com escribió:
Hi David,
I think it will work quite well to just use the same data element in different datasets with different periodicity. The only thing to look out for is to avoid double counting, e.g. watch out for misleading Totals in Analytics.
It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.
My instincts said two DE’s as well … but could see you could possibly ‘get-away’ with a ring-fenced data-entry strategy for one data-element in two datasets. We’ll do a little bit more experimentation with a sample of historical and current data - we’re just trying to avoid complicating our custom reports and analytics for users of the data.
22.2.5. Data elements assigned to data sets with different period types
Data elements should not be assigned to two separate data sets whose period types differ. The recommended approach would be to create two separate data elements (for instance a monthly and yearly data element) and assign these to respective datasets.
I am sure Lars can provide the exact reason for this, but I am pretty sure one should not have the same data element with two separate frequencies. It would seem to be a good feature however, but I believe this is a restriction imposed by analytics at the moment.
Regards,
Jason
On Thu, Jul 14, 2016 at 4:28 PM, Knut Staring knutst@gmail.com wrote:
Thanks JM, I forgot to mention that.
Though you may still have to watch the totals if you combine months and years in the same Pivot Table.
Hi Knut
One way to avoid double counting is to make sure both data sets are never active in the same org units.
JM
El 14/07/2016, a las 2:10 p.m., Knut Staring knutst@gmail.com escribió:
Hi David,
I think it will work quite well to just use the same data element in different datasets with different periodicity. The only thing to look out for is to avoid double counting, e.g. watch out for misleading Totals in Analytics.
It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.
I think I have on a couple of occasion imported legacy data with a different period type than what is currently collected without any problems with the analytics.
22.2.5. Data elements assigned to data sets with different period types
Data elements should not be assigned to two separate data sets whose period types differ. The recommended approach would be to create two separate data elements (for instance a monthly and yearly data element) and assign these to respective datasets.
I am sure Lars can provide the exact reason for this, but I am pretty sure one should not have the same data element with two separate frequencies. It would seem to be a good feature however, but I believe this is a restriction imposed by analytics at the moment.
Regards,
Jason
On Thu, Jul 14, 2016 at 4:28 PM, Knut Staring knutst@gmail.com wrote:
Thanks JM, I forgot to mention that.
Though you may still have to watch the totals if you combine months and years in the same Pivot Table.
Hi Knut
One way to avoid double counting is to make sure both data sets are never active in the same org units.
JM
El 14/07/2016, a las 2:10 p.m., Knut Staring knutst@gmail.com escribió:
Hi David,
I think it will work quite well to just use the same data element in different datasets with different periodicity. The only thing to look out for is to avoid double counting, e.g. watch out for misleading Totals in Analytics.
It’s been a while … changing mail clients meant that forum messages skipped my in-box … and what do they say … ‘out of sight - out of mind’.
Anyway, I’ve rearranged settings and can see the email list is as busy as ever!
And onto my burning question:
If we have a mix of organisational units (countries in our case, collecting national data) that:
some of which have historical monthly data, and some of which only have annual data; and
going forward, are transitioning to monthly data (but some will be slower and for the next couple of years will only have annual data)
AND we want to bring this into a single DHIS2 instance, then
Is the best approach to create a set of duplicate data elements (in different datasets, one set for an annual collection period and one set for a monthly collection period)? or
Is it possible to use a single data element in two different datasets with different periods set (one monthly and one annual)? … (I don’t see how this would work in practice, but maybe I’m just missing something - I know it is possible to put a DE in more than one dataset, but mixed collection periods for the same data element seems like it will mess up reporting and analytics)
Any thoughts or guidance or other options from the group appreciated.