search

zazuko / elcom-strompreis-poc

PoC ElCom Strompreise LD
0 stars 0 forks source link

Feedback/questions from the Map PoC #4

Open jstcki opened 4 years ago

jstcki commented 4 years ago

Hi!

We made progress with the municipality map PoC -> https://elcom-poc.netlify.com/en/map (I'll send you the PW separately) and now have some questions about the data:

  1. There is only one attribute VSE which we're not sure what to do with http://yasgui.org/short/kOsg60fSh
  2. Probably expected but there's no unit attributes for the measures
  3. When compared to the original Strompreis website, some numbers are off. E.g. for "Safiental" http://yasgui.org/short/2vkculGw6 cf. https://www.strompreis.elcom.admin.ch/Map/ShowComMap.aspx?CatID=4&CantonBez=GR&View=0&Period=2019&ProdID=10 and https://www.strompreis.elcom.admin.ch/PriceDetail.aspx?Period=2019&PlaceNumber=3672&OpID=342&CatID=4&ProdID=10&CPeriod=2019&CPlaceNumber=3672&COpID=662&CCatID=4&CProdID=10&DetailData=Comp
  4. Municipality 4238 seems to be an outlier with KEV 0 ? But everything else is always KEV 2.3 http://yasgui.org/short/HzIHun9V_

/cc @bldng

jstcki commented 4 years ago
  1. There seems to be a "Product" dimension on the original site that is not available in "our" data.
  2. Just FYI: Some of the labels have typos ("Engergy" etc.) or are not translated ("reference year")
ktk commented 4 years ago
ktk commented 4 years ago

1: Regarding VSE: I don't know much more that's why I attached it as attribute. In the Excel it's simply called "VSE-ID" but I could not figure out more than that.

ktk commented 4 years ago

3: @herrstucki If I query this manually for provider Repower I get the same data like on the Elcom website. I can't open the SPARQL query, I get a timeout from this link. Will check later again

ktk commented 4 years ago

5: There is no such dimension in the available raw data it seems.

bldng commented 4 years ago

Hi!

3: Attached is the difference to the elcom website.

The mean aggregation happens on the map view, but the table view shows the same values as the SPARQL query. All requests should show values by H4 for 2019. At first we supsected a 1-year shift because the prices refer to the coming year, but that does not seem to be the case.

image

ktk commented 4 years ago

4: I checked that and found a bunch of providers that have KEV as 0, see http://yasgui.org/short/jks5Iv154. I went through a bunch of them in the original data and can confirm that they indeed have 0 for KEV so I would say that is correct.