fxedel / vaterstetten-in-zahlen Goto Github PK

Data update dd2a9eb fetched a photovoltaic system for a single family home with 4.6 megawatts. This seems implausible, so probably a plausibility check should be introduced for that.

Steps to reproduce:

1. Go to https://vaterstetten-in-zahlen.de/?tab=kommunalwahl2020 (directly or via an in-app link)
2. See this result:
   
3. See these console errors:
   

Changing any dropdown loads the map data as expected, also changing it back to the default value works.

Most German statistical offices use GENESIS as public database, e.g.:

• https://www-genesis.destatis.de/genesis/online
• https://www.statistikdaten.bayern.de/genesis/online

These databases offer a lot of interesting data about Vaterstetten. For example, we use them for the einwohner dataset.

To regularly update the data and make the data generating process more transparent, a new poller for this dataset is more suitable than importing the data by hand. It should be reusable, to use it for different datasets.

In this plot, highlighting the 50% line with a dark grey dashed line (or similar) might help comparing the numbers.

Right now, the three value boxes on the vaccination page display the following numbers:

1. Absolute number of vaccinated people (first vaccination), e.g. 64231
2. Percentage of vaccinated people (first vaccination), e.g. 44.7%
3. 7-day impfidenz, e.g. 5526.7

I suggest to change this:

1. Absolute number + percentage of vaccinated people (first vaccination), e.g. 64 231 (44.7%), user-clock icon
2. Absolute number + percentage of vaccinated people (second vaccination), user-check icon
3. 7-day impfidenz, e.g. 5526.7

Rationale: There is now easy way to read the percentage of fully vaccinated people from the graphs below.

We already gather batteries from the Marktstammdatenregister, so we should also visualize this data.

Some things to keep in mind:

• Do data cleaning and plausibility checks; people tend to mix up power and capacity
• Somehow count how many batteries have been installed together with a photovoltaic system?
• Plot different battery types (lithium, redox-reflow, ...)

With this Overpass query, you can get all bench nodes in Vaterstetten in OpenStreetMap:

[out:json];
area
  ["boundary"="administrative"]
  ["name"="Vaterstetten"]->.a;   // Redirect result to ".a"
out body qt;
(
  node
    (area.a)                     // Use result from ".a"
    ["amenity"="bench"];
);
out body qt;
>;
out skel qt;

You can also query historical data (since 2012-09-12T06:55:00Z) using the [date:"2015-10-28T19:20:00Z"] setting, see https://overpass-turbo.eu/s/1ilC.

It would be interesting to see how the number of benches in Vaterstetten (in OSM) grew over time, maybe in a monthly interval.

Maybe there also other interesting tags one could plot in a similar manner.

The Landkreis Ebersberg biyearly publishes a greenhouse gas balance:

• 2016: https://www.energieagentur-ebe-m.de/data/dokumente/konzepte%20ebe%20m/2020_05_12_Treibhausgasbericht_Landkreis_Ebersberg_2016_komprimiert.pdf
• 2018: https://www.energieagentur-ebe-m.de/data/dokumente/konzepte%20ebe%20m/2020-12-18_Treibhausgasbericht_Landkreis_Ebersberg_2018_web.pdf
• 2020: https://www.energieagentur-ebe-m.de/data/dokumente/sonstige/2022_11_22_THG_Bilanzierung_final_Web.pdf

The report also includes data aggregated by municipality, so we do also have numbers specifically for Vaterstetten.

As of March 2021, it was not intended to publish the data in a machine-readable format, but it could be worth asking if this would be possible for the 2020 report (published in 2022).

The Bayernwerk Netz GmbH has an energy dashboard for the Landkreis Ebersberg: https://energiemonitor.bayernwerk.de/ebersberg-landkreis

And they have an API which gives us the raw data:

• https://api-energiemonitor.eon.com/tenants
• https://api-energiemonitor.eon.com/region-data?regionUrlKey=ebersberg-landkreis
• https://api-energiemonitor.eon.com/meter-data?regionCode=09175
• https://api-energiemonitor.eon.com/historic-data?regionCode=09175
• https://api-energiemonitor.eon.com/weather-data?regionCode=09175&tenantId=3190

The meter-data gives us the current values in a 15-minute resolution, the historic-data gives us historic values in a 24-hour resolution. The latter seems to be more appropriate to be polled. It would also be interesting having the values in a 15-minute resolution for every day, but this would require us to very reliably poll the meter-data endpoint every 15 minutes.

The visualization is already quite good, we won't need to copy everything. However, the history data for the energy consumption per cluster (domestic, industrial, public) is missing.

Originally suggested in #44 (comment):

In the new street name page, it could be displayed which street name prefixes/suffixes are most common. A bar plot probably makes most sense, although a map would also be nice. The list to check would need to be hardcoded (just like the categories are):

• "-straße" (currently 131)
• "-weg" (currently 113)
• "-ring" (currently 6)
• "-platz" (currently 5)
• "-anger" (currently 4, including "Am Anger")
• "-hof" (currently 2)
• "-gasse" (currently 1)
• "-allee" (currently 1)
• "-spange" (currently 1)
• "-winkel" (currently 1)
• "Am …" (currently 9)

• Number of vehicles by municipality (Kommune): https://www.kba.de/DE/Statistik/Produktkatalog/produkte/Fahrzeuge/fz3_b_uebersicht.html?nn=3514348
• Maybe there are other statistics by municipality? Very interesting would be the number of electric cars. -> Maybe also try to get data from the Zulassungsstelle (e.g. via email)

Similar to https://github.com/EqualStreetNames/equalstreetnames/ (see e.g. https://munich.equalstreetnames.eu/), visualize the streets that are named after persons and calculate the percentage of male and female eponyms.

They use OpenStreetMap data and its name:etymology:wikidata tag.

Right now, you can select the date via the dateRangeInput, which is quite cumbersome though.
E.g. to view all Corona cases (which only displays 4 weeks by default), I have to click 13 times on "previous month".

I suggest adding the following buttons to select the date range:

• last 7 days / last week
• last 30 days / last month
• last 90 days / last 3 months
• last 365 days / last year
• all data

The UI could then look like this:

Daten anzeigen von [12. Mai 2021 bis 9. Juni 2021] [letzter Woche] [letztem Monat] [letzten 90 Tagen] [letztem Jahr] [gesamten Zeitraum]
☑ Zahlenwerte anzeigen

The button "last year" doesn't make sense for vaccinations right now, because the there isn't enough data, so it shouldn't be displayed there.

Right now, only 7-day incidence and 7-day impfidenz are displayed, which is a good default, because those are known and used everywhere.

It might be interesting to see the same data smoothed over a longer time period though. I suggest to add a dropdown (right there in the tile containing those visualizations, e.g. next to the title) where you can choose between 7 days (default) and 14 days. The title and description in the tile have to change accordingly of course.

While we already have the data, we could visualize more interesting features:

• orientation and angle? -> this could also be interesting if we had numbers of the energy production (which should be possible for EEG installations), to show how effective are modules on the north side, etc.
• self-use vs. network feed
• map with biggest installations
• table with biggest/known installations
• table with special installations: Mieterstrom, EEG-Ausschreibung
• pie charts about total distribution by location, building type, installation size