Issue with series_parsers and PT4S

[Xero57]

Hello,

I was unable to access Cross Border Marginal Prices for aFRR (aFRR IF 3.16) due to resolution being 4 seconds (PT4S) which is not defined in _resolution_to_timedelta in series_parser.py when using the EntsoePandasClient. The 4-second resolution is relatively new and used only for aFRR as far as I am aware.

Thank you for the great API client.

[fboerman]

please post a code snippit as well as a link to the transparancy platform of which data it is. Thanks!

[Xero57]

please post a code snippit as well as a link to the transparancy platform of which data it is. Thanks!

Here's the code:

client = EntsoePandasClient(api_key=api_key)

start = pd.Timestamp('202301010000', tz='Europe/Brussels')
end = pd.Timestamp('202301020000', tz='Europe/Brussels')
country_code = 'CZ'
process_type = 'A67'


ts = client.query_activated_balancing_energy_prices(country_code, 
                                                    process_type=process_type,
                                                    start=start, 
                                                    end=end)
ts.to_csv('outfile.csv')

Here is the link for the API documentation: https://documenter.getpostman.com/view/7009892/2s93JtP3F6#d8e9ac28-1687-4339-bc18-7afad2bf028c

Here is the link for the data view for the data I was looking for: https://transparency.entsoe.eu/balancing/r3/cbmpsForAfrrStandardProduct/show

Here is the article for the data on the API knowledge base: https://transparencyplatform.zendesk.com/hc/en-us/articles/12826703714452-Energy-CBMPs-for-aFRR-standard-product-aFRR-IF-3-16

I am an amateur programmer, so it is possible I simply did not understand some part of the code correctly or made a mistake and it would not work anyway for the data I wanted. Aside from that, I still could not find where the PT4S resolution type is in the code, as I did not find it hard coded with the other resolution types. So I opened an issue as was advised in the NotImplementedError.

[fboerman]

hi @Xero57 it seems that the response generated by this specific process type is very different from the other things. this required more custom code to deal, I dont have that much time available now.

if you want you can check the _parse_activated_balancing_energy_prices_timeseries function, put a breakpoint there and check the output that is being parsed in the for loop on what to adjust. if you succeed you can open a pull request and ill merge it.