Concatenated file questions + numbers rounding + table row names in content

[marfi]

Hi,

Admirable effort, congratulations!

I checked the .pdfs, however in the concatenated file, I see these fields which are not described in the "dataset_descriptions.pdf". The bolded are the most important ones for me:

project_name,
beneficiary_name,

total_amount,

eu_cofinancing_amount,
total_amount_eligible,
eu_cofinancing_amount_eligible,

amount,

amount_kind,
approval_date,
beneficiary_address,
beneficiary_city,
beneficiary_country,
beneficiary_country_code,
beneficiary_county,

beneficiary_id, with whom? local authority, openspending ?

beneficiary_nuts_code,
beneficiary_nuts_region,
beneficiary_person,
beneficiary_postal_code,
beneficiary_url,
cci_program_code,
completion_date,
currency,
eu_cofinancing_rate,
final_payment_date,
first_payment_date,
fund_acronym,
funding_period,
management_authority,
member_state_amount,
operational_programme,
priority_label,
priority_number,

project_description,

project_status,
source,
starting_date,
theme_code,
theme_name,
third_party_amount,
total_amount_applied

• In general, the concatenated file has much more fields than the pdf
• the project_id is missing from the concatenated file :(
• what's the difference between the total_amount and amount?
• why is project description so far away from the project name?

Also, the very first line, contains this number:

898269689.3997617683926989018 in the: amount row (7).

I think it should also be rounded to .nn, like the total_amount (3) and the eu_cofinancing_amount (4). I can do the rounding on my end, but stil...

Row 8 contains this: "eu_cofinancing_amount", the header is repeated if the column is empty. I can clean this, but still...

[michaeljokf]

@marfi thanks for your feedback! What you looked at is the dataset descriptions, only a brief description of what the original files were when we found them. You should read this methodology. Therein all the variables the datasets include are discussed.

If you really want to work with the datasets you should assign unique ids!
Project ids were neglected, because they were come from the original datasets (were assigned by the local administrations) and do not offer value in the concatenated data.

We will look into the large (not rounded) number and the repetition of the "eu_cofinancing_amount" title.
If your questions remain after looking at the methodology, feel free to comment again, it does go into detail on all the variables though.

[michaeljokf]

@marfi about "beneficiary_id" we added this rather last minute, it was created using https://github.com/pudo/fingerprints to assure that beneficiaries are actually spelled the same. While this does not necessarily look nice, it allows grouping the beneficiaries correctly. The results from this are helpful for these countries: Austria, Estonia, Germany, Italy, Netherlands, United Kingdom.