Closed sarahshamiso closed 1 year ago
Initial contact email sent to members in registry for BRC.
BRC response: Hi David,
Thanks for your email.
I understand the concern with the data and let me reassure you and the team that we identified this as cause by a system error and our developers are already working on a solution. I actually submitted the data two weeks ago and the system fix on our information system where data are collected, will be available in the coming days. However, I should be able to upload a correct data export on the IATI website by the end of the week as the developers have committed to provide me with a correct data export before the solution is implemented on our site.
Please let me know if you have any question.
Hi @sarahshamiso @DaveEade49 thanks for looking into this
Regarding the specific COVID activity - I can still see that some transactions appear to be rather large - eg this one for £35m
When you first identified the issue, can you recall if this was in millions or trillions?
Furthermore, I'm struggling to understand why this is a Disbursement for this activity, when it could / should actually represent an Incoming Funds from FCDO to IFRC (published by BRC)...?
In terms of the IATI COVID dashboard, we've temporarily suspended use of this activity
Hi @stevieflow
I suspect this was billions before -- in my myriad of links above I shared the data from 4 Oct and also sharing it here with a tab I've added for this particular activity. I haven't re-run the data but the values are much smaller than they were. Before it showed that FCDO gave 47 billion to IFRC in Q2 2020 and now it shows about 36 million with 2 transactions. I can't say whether this is now correct but looks better? :)
British Red Cross data_downloaded 4Oct2022.xlsx
I will re-run all of the data later in the week to check whether they appear to have fixed things across the board.
Regarding the flow of funds, great question and I'm not sure what is going on here (but agree this seems to be incorrectly recorded or an odd use of the disbursement transaction) -- would need to investigate further but perhaps @DaveEade49 can look into it.
I've passed the query regarding @stevieflow observation to BRC today - hopefully they will review and get back to us. My understanding is that BRC folks are new to this so may need a bit of hand holding or advice - if that is the case @stevieflow are you available to support? Thanks Dave
Thanks @sarahshamiso @DaveEade49
That's really great to hear - hopefully we can resolve this
In terms of the various purpose of the flows and how they might best fit, I can try and help, of course. @pelleaardema may also be interested, too...
BRC have confirmed that the 35.9m figure is correct, my contact has referred me to other members on the BRC team to answer the more detailed question you raised @stevieflow
Using the data behind IATI data dump, you can browse old versions of this data: https://gitlab.com/codeforIATI/iati-data/-/commits/main/data/gb-chc-220949/gb-chc-220949-activities.xml
If you click the little folder icons, you can see the data as it stood after each update.
This shows the data was updated three weeks ago, at which point lots of transaction amounts became massively inflated. E.g. £35,900,000.00 became £35,935,900,000. This example looks fixed in the update 3 days ago.
can we close this one off now? @sarahshamiso
Hello! There has been no activity on this issue in the last 30 days. I wonder if it has now been resolved?
If you’re reading this, would you mind checking to see if the issue is still applicable?
Thank you!
can we close this one off now? @sarahshamiso
Agreed – In fact it looks like this was resolved by the 12th October. E.g. this transaction is back to £35,900,000.00.
Thanks all! 🙏🏻
The British Red Cross seem to have made some major errors in their disbursement transaction values, resulting in aggregate spending of 273 trillion in 2020, 41 trillion in 2021, and 3 trillion in 2022. File was last updated end of August 2022.
Looking at their data, I suspect they have just put the decimal place in the wrong spot or added too many zeroes.
I've attached the file from CDFD that combines all data for British Red Cross where you can clearly see the issue in the first two tabs.
An example activity is also here: http://d-portal.org/ctrack.html#view=act&aid=GB-CHC-220949-P8092 with XML here: http://d-portal.org/q.xml?aid=GB-CHC-220949-P8092
Have alerted the technical team to the issue.
Also will flag a related issue on d-portal -- in researching this issue, discovered that some activities on d-portal don't contain this more recent data because d-portal only includes the first instance of an activity identifier and the identifier appears in the file containing historical data and also in the file with the updated data so the new transactions from this newer file are missing. But oddly, the Registry shows that the file from which d-portal is pulling the data is not an IATI file -- so unclear how d-portal is actually pulling in this data. Will raise separately on the d-portal github.
Registry files: https://iatiregistry.org/publisher/gb-chc-220949
Validator showing file issue: https://iativalidator.iatistandard.org/view/dqf/files/455d478f-b046-4e4f-a809-b10e2567e428
Example d-portal activity: http://d-portal.org/ctrack.html#view=act&aid=GB-CHC-220949-P7803
File with data: British Red Cross data_downloaded 4Oct2022.xlsx