rastrau / monitoring-geodienste-ch

Analysis of geodata offerings of Swiss cantons on the geodienste.ch platform
https://ralphstraumann.ch/monitoring-geodienste-ch
5 stars 0 forks source link

new datasets #6

Closed romanmeyerblch closed 1 week ago

romanmeyerblch commented 2 weeks ago

I'm very happy to see that this website is still up, running and being developed. As far as I understand, additional services/datasets are not automatically included in the analysis, e.g. Gewässerraum. Would be great if those could also be included, otherwise the overall picture will slowly but surely become incomplete.

rastrau commented 1 week ago

Thank you, @romanmeyerblch! As a matter of fact, new / additional datasets should be included. You can see this e.g. in the "unbekannt" datasets that go into the data, in Fig. 1:

image

Specifically for Figs. 1, 2 and 3 and the hover texts that are displayed, the analysis uses (inofficial, i.e. made up by me) abbreviations for long dataset names. The long names themselves can't be displayed properly. An example for such a substitution would be: "BdF" instead of "Biodiversitätsförderflächen, Qualitätsstufe II und Vernetzung". This is done in the code e.g. here.

Since, short of employing some AI, I cannot dynamically define sensible and unique abbreviations a priori, the code substitutes "unbekannt" for dataset names that it doesn't know yet. Even with the "unbekannt" monicker these datasets and their availability and openness still enter the analysis and the results. Over the previous months, some datasets turned to "unbekannt", simply because KGK-CGC have renamed them (several focussed on agriculture, e.g.). Besides, it seems "Gewässerraum" ("Gwr") and "Rodungen und Rodungsersatz" ("RuR") are datasets that have genuinely been added since the last update of the monitoring code.

I have updated the set of abbreviations (https://github.com/rastrau/monitoring-geodienste-ch/commit/9b9e72ed074f4a5adaf7ed40325a44945741dc0f, https://github.com/rastrau/monitoring-geodienste-ch/commit/c58835b2a83eaaf21119f15c297a2efb86376e59) and re-run and re-rendered the analysis. You shouldn't find any "unbekannt" datasets anymore.

Please let me know if you think anything of what I commented is wrong/implausible or if you see any other areas for improvement. Also always happy to collaborate (albeit the code might not be ideally structured for that yet).

romanmeyerblch commented 1 week ago

Great, thanks for the explanations @rastrau!