Thanks for looking at this and getting back to us.
···
On Thu, May 21, 2015 at 8:31 PM, Tran Chau (HISP Vietnam) tran.hispvietnam@gmail.com wrote:
Hi Shantala,
I took a look at your database and run it on 2.19 version.
It seems that you want to overwrite some data in aggregate side (dataSet values) by removing events on event side and running the manual aggregation, correct?
If so, DHIS ‘aggregation’ from event side does not work like that.
If there is no events, the aggregation process simply skips aggregation for that. It does not write over ‘0’ values to aggregation side (dataSet values).
…
Best regards,
Tran Chau
On Tue, May 19, 2015 at 9:42 AM, Tran Chau (HISP Vietnam) tran.hispvietnam@gmail.com wrote:
HI Shantala,
I will take a look at it couple days later. I will inform you if I find out the issue.
…
Best regards,
Tran Chau
On Mon, May 18, 2015 at 7:54 PM, Shantala Raman sraman@thoughtworks.com wrote:
Hi Tran,
Here is the link to the DB dump of DHIS in dropbox where we have setup the data needed to replicate the defect. We have associated the Cholera CTC program under the organization hierarchy, MSF->OCP->A new country->A new project->New Hospital->Cholera->Origin 1. The program is of type Single event without registration.
Steps to reproduce the defect:
- In event capture, register an event for the Cholera CTC program under the Origin 1 org unit.
- Preview the manual aggregation results for Geographic Origin data set for Origin 1 org unit for the week of the Incident date and save all values.
- In data entry, you will see that the value of the Number of Patients (Cholera CTC) data element for the Geographic Origin data set for the week of the Incident date is 1.
- In event capture, remove the event for the Cholera CTC program under the Origin 1 org unit.
- Preview the manual aggregation results for Geographic Origin data set for Origin 1 org unit for the week of the Incident date. This won’t return any changes to be updated while the expected behavior is that preview shows that the value for the Number of Patients (Cholera CTC) data element will change from 1 to 0. Save all values.
- In data entry, you will see the value of the Number of Patients (Cholera CTC) data element remains as 1 for the Geographic Origin data set for the week while the expected behavior is that the value of the Number of Patients (Cholera CTC) data element is 0.
Thanks,
Shantala
On Tue, May 12, 2015 at 3:05 PM, Shantala Raman sraman@thoughtworks.com wrote:
Hi Tran,
We are on 2.19 rev 19092. We are not using SQL queries for generating summary data values. We are using the aggregation query builder to setup the queries. Could you explain how the SQL query is linked to aggregation query builder?
Thanks,
Shantala
On Mon, May 11, 2015 at 7:10 PM, Tran Chau (HISP Vietnam) tran.hispvietnam@gmail.com wrote:
Hi Shantala,
The first, please make sure if the snapshots are from 2.19 version, R 19075 ? It looks like old version.
I couldn’t create an example on dev server ( apps.dhis2.org/dev ) because the internet is still slow here. I cannot login the dev server to create an example.
I am sending the SQL statement which I used to generate the data values from aggregate query formulas.
Can you change the IDs in the SQL statement below according to one of your formula and see if it is generated “NULL” value records or not ( these NULL values will be displayed/saved as “0” values in Manual Aggregate )
SELECT ‘55’ as dataelementid, ‘16’ as categoryoptioncomboid,
‘16’ as attributeoptioncomboid, ou.organisationunitid as sourceid,
‘1234’ as periodid,‘aggregated_from_tracker’ as storedby,
‘201505’ as periodIsoDate,‘Agg DE 1’ as dataelementname,
‘(default)’ as categoryoptioncomboname, ou.name as organisationunitname,
( select count(psi.programinstanceid) as value FROM programinstance as pi
INNER JOIN programstageinstance psi ON pi.programinstanceid=psi.programinstanceid
WHERE EXISTS ( SELECT * FROM trackedentitydatavalue _pdv
inner join programstageinstance _psi ON _pdv.programstageinstanceid=_psi.programstageinstanceid
JOIN programinstance _pi ON _pi.programinstanceid=_psi.programinstanceid
WHERE psi.programstageinstanceid=_pdv.programstageinstanceid
AND _pdv.dataelementid=53 AND _psi.organisationunitid in (50, 51, 52)
AND _pi.programid = 60 AND _psi.executionDate>=‘2015-05-01’
AND _psi.executionDate <= ‘2015-05-31’
AND _psi.programstageid = 63 AND _pdv.value is not null )
and psi.organisationunitid=ou.organisationunitid
GROUP BY ou.organisationunitid ) from organisationunit ou where ou.organisationunitid in ( 50, 51, 52 )
…
Best regards
Tran Chau
On Mon, May 11, 2015 at 8:02 PM, Tran Chau (HISP Vietnam) tran.hispvietnam@gmail.com wrote:
Hi Shantala,
Are you sure if the snapshots are from 2.19 version, R 19075 ?
It looks like old version.
…
Best regards,
Tran Chau
Hi Tran,
We are still seeing issues with the aggregation. We followed the below steps -
- Created a data element with Store zero data values property set to Yes
- Created an aggregation query for the data element
- Created 3 events for the program of type single event without registration - 2 events under orgunit 1 (origin 1) and 1 event under orgunit 2 (origin 2)
- Ran manual aggregation for both org units
- Deleted the event under org unit 2 (origin 2)
- Ran manual aggregation for both org units
We observe that the data entry screen for org unit 2 (origin 2) still has the number of events as 1.
Please find attached the screenshots for the various steps. Please let us know if you need any other information.
Thanks,
Shantala & Shashank
On Fri, May 8, 2015 at 8:12 AM, Tran Chau (HISP Vietnam) tran.hispvietnam@gmail.com wrote:
HI Simata,
Yes, this bug is fixed on 19075, trunk version.
I created an example on dev server ( https://apps.dhis2.org/dev/ ) for testing.
I am sending snapshot about the test step by step on dev server here.
Step 1 - Create a data element and an aggregate query builder formula
Step 1.1 - Create a data element with*'*Store zero data values’ property as 'Yes’ ( see snapshot [1.1] ). And add it into a dataset.
Step 1.2 - Defined an aggregate query builder formula*( see snapshot [1.2] )*
Step 2 - Generate data values
Step 2.1 - Create a new person and an event in April 2015. Fill a data value for data element which I used in Step 1.2 ( see snapshot [2.1] )
Step 2.2 - Run Aggregate Manual ( see snapshot [2.2] )
Step 2.3 - Result after clicking “Preview” button ( see snapshot [2.3] )
Step 2.4 - Data values imported in the data set ( see snapshot [2.4] , [2.5] )
Step 3 - Modified data value of event in Step 2.1 and see the result after aggregate
Step 3.1 - Remove the data which was filled in Step 2.1 ( see snapshot [3.1] )
*Step 32 *- Run Aggregate Manual ( see snapshot [3.2] )
*Step 3.3 *- Result after clicking “Preview” button ( see snapshot [3.3] )
Step 3.4 - Data values imported in the data set ( see snapshot [3.4] , [3.5] ).
We can see the data value in April ( snapshot [3.4] ) is 0 now )
…
If you got different result, send me the images and description like these.
We will take a look at it.
…
Best regards,
Tran Chau.
On May 7, 2015 1:03 PM, “Simanta Jyoti Handique” shandique@thoughtworks.com wrote:
Hello,
We were wondering if this bug has been fixed as part of rev #19075. We are on rev #19097 and are still facing this issue
Regards,
Simanta
On Tue, Apr 28, 2015 at 5:51 PM, Tran Chau (HISP Vietnam) tran.hispvietnam@gmail.com wrote:
HI Simanta and others,
The issue is in Manual Aggregate, the “0” values are not displayed/saved.
I am working on it to display and save “0” values in Manual Aggregate functionality.
Thank you for reporting this bug.
…
Best regards,
Tran Chau
On Mon, Apr 27, 2015 at 5:48 PM, Simanta Jyoti Handique shandique@thoughtworks.com wrote:
Hi, thanks for the response. The data elements were already configured to store zero data values.
This may be another symptom to the problem: If I run the aggregation manually and preview the data that will be saved, there are no results with 0 values
Regards,
Simanta
Mailing list: https://launchpad.net/~dhis2-devs
Post to : dhis2-devs@lists.launchpad.net
Unsubscribe : https://launchpad.net/~dhis2-devs
More help : https://help.launchpad.net/ListHelp
On Mon, Apr 27, 2015 at 1:58 PM, Adebusoye Anifalaje busoye@hisp.org wrote:
Sorry,
Just re read your question. My mistake. To overwrite values with 0, you have to choose to store zero data values when creating the aggregate data element.
On Monday, April 27, 2015, Adebusoye Anifalaje busoye@hisp.org wrote:
Hi Simanta,
You should try changing the Operator to Number of tracked entities. Visits are technically for program stages and since you are not registering programs then you can’t have visits.
Try tracked entities and let’s know whether it worked.
Cheers.
Busoye
On Monday, April 27, 2015, Simanta Jyoti Handique shandique@thoughtworks.com wrote:
Hi,
We have set up an aggregation query with the following details:
Name: Number of Patients (Measles Outbreak)
Operator: Number of Visits
Data set: Geographic Origin
Data Element: Number of Patients (Measles Outbreak) - Geographic Origin (Number)
Condition: [Measles Outbreak.Measles Outbreak stage.Date of admission] is not null
“Measles Outbreak” is a Single Event without Registration type program. “Date of admission” is a Data Element of type Date
We register an event, run the aggregation query and the “Number of Patients (Measles Outbreak) - Geographic Origin (Number)” data element is correctly populated with the count of the number of patients. We delete the event and re-run the query. The data element for the number of patients does not get updated with the value 0.
Is this a defect with the Aggregation Query feature or is there something wrong with the formula we’re using?
Regards,
Simanta
