Versions Compared

Version Old Version 4 New Version Current
Changes made by

Ben Leibert

Dércio Duvane

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  • It would be great to confirm a number of things related to the Data Sources page. (Doing so may potentially take a while.)

  • We decided to start using Jira to tack our tracks, much like we do with VR's other projects.

  • Review the status of Action Items assigned at past meetings.

Discussion Items

Data Sources - SELV

OpenLMIS treats "Sarampo" and "Sarampo 2 MSD" as two separate antigens. It really shouldn't though - they both refer to an antigen called "MR" (which stands for Measles and Rubella.) Although the SELV+ Dashboard currently references "Sarampo," it should instead just list "MR" (which, again, should represent a combination of OpenLMIS' "Sarampo" and "Sarampo 2 MSD".

The existing Excel dashboard includes the province of “Inhambane” on the Dados_Stock sheet. This is an experimental province which does not exist in OpenLMIS, and which need not exist in future versions of the SELV+ dashboard until it's introduced into OpenLMIS.

OpenLMIS invites users to enter "Number of Stockout Days." Why, then, do most of the rows in Excel's Indicator List related to SELV say "SELV will give us the stocks. A separately calculation will be necessary to determine stock out?" We discussed that this may be because the MoH potentially wants to define 1 "stockout" event within the SELV+ Dashboard as 5 consecutive days without stock. Is this different from what users are entering into OpenLMIS, though? Depending on how people use this field in OpenLMIS, and because OpenLMIS exclusively collects data with a monthly level of granularity, it may not be possible to produce statistics that are congruent with the MoH's definition of stockout. We need to better understand what they want, and Dércio Duvane will help us.

Clay's notes on SELV data:

  • PCV
    • PCV → is now PCV13
    • Some provinces still using PCV10 → so we should use
    • Refer to as PCV on dashboard
  • Sarampo
    • Sarampo and Sarampo 2 are the same antigen, just for different doses
    • Call both Sarampo and Sarampo 2 “MR” in SELV dashboard
  • Stockout days
    • Team in Mozambique is confirming definition of stockout



Data Sources - DHIS2

We have no need to directly access the DHIS-2 database. Instead, we should be able to pull the data we need from their public API using the same set of credentials they've already given us (and which appear on the /wiki/spaces/SELVDashboard/pages/365756559 page. Nevertheless, to avoid any surprises, Dércio Duvane will potentially let the MoH know that we plan to use their API. He'll emphasize, though, that we'd only reading data - and that we're certainly not accessing their database directly.

Craig Appl (Unlicensed) is currently blocked. He needs to determine exactly how we should retrieve data from the DHIS2 API. To do so, however, he needs to understand exactly what data we're currently pulling from the system. To convey this, Craig needs annotated screenshots of the DHIS-2 user interface which depict where users currently go to extract data from the system. Please indicate along with the screenshots 1) What you're clicking, 2) What report the system subsequently returns, and 3) exactly how the fields within the DHIS-2 report relate to the raw-data columns within the existing SELV dashboard. Dércio Duvane will help provide this info and will ask for clarification about the request if need be.

Data Sources - HSS Funds 

...

By default, SELV+ Dashboard users should only see the data represented by column E of the Excel spreadsheet. The data in columns B, C, and D should be accessible only to users with an elevated permission level.

Clay's notes on processing HSS Funds data and calculating indicators:

  • Use top table - 134GAVI - HSS
    • The rest should be discarded
    • Second and third table are just for specific projects that are concluded - we don’t need to import on a regular basis
  • Each row is a province - data is not tracked at district or facility level
    • Top row is the MoH value (should be treated as a province)
      • We will create a row each month that is null for all values except month, province, starting HSS balance (col B), HSS funds spent this month (calculation made by comparing col C from month-to-month)
      • We will calculate cols D and E in Superset
    • Row 19 is national level
    • Col B is grant total at beginning of year
    • Col C is execution of expenses (what is spent) from beginning of year to date
    • Col D is remaining balane
    • Col E is execution rate (C / B * 100)
  • Getting monthly-level numbers, to compare burn rates across months:
    • We will load the raw data into the first (source-specific) table
    • We will calculate the expenditures of the current month by subtracting the YTD total for the previous month from the YTD total from the current month 



Data Sources - NexLeaf

Although we'd like to pull raw data from NexLeaf via an API, the NexLeaf team doesn't have time to provide one. We've thus been planning to rely on imagines that we can pull from URLs they currently offer. The drawback to this approach, however, is that the SELV+ Dashboard would loose the ability to show the following pieces of data:

...

If we do decide to rely on pulling images, we hope to avoid the performance penalty of HTTP requests to NexLeaf by retrieving the images as part of the data-ingestion process. The images would thus essentially be cached within the local database. Clay Crosby (Unlicensed) and the Ona team think this approach would probably work.

Finally, the section of the indicators list shown below would benefit from elaboration. Dércio Duvane will thus review and update it.

Image Added


Data Sources - Trimestrial Reports

...

Dércio Duvane: Please feel welcome to assign items to Timoteo Chaluco (Unlicensed) as the two of you see fit.