Stats & graphs

[raphodn]

Ideas

model	detail	example
Product	Number of prices Number of locations/countries Map of locations (with prices)
Product	Evolution of price
Category	min, max, median price
Product	Evolution of price per location/store chain
Location	Number of prices Number of different products Number of different users
Location : city	Number of cities with at least 1 price Cities with the most locations Cities with the most prices
Location : country	Number of countries with at least 1 price Countries with the most locations Countries with the most prices
Location : type	Number of types with at least 1 price Number of prices per type	`
Location : name	Number of prices per store name/chain
User	Number of prices Number of different products Number of different locations
Proof	Number of prices per type

filters on date, location store, location country, user, category...

Related issues

• Product stats endpoint #254

Related discussions

• Gamification #399

[serpico]

Related to #302 @raphodn could you pull some stats ( similar to what you did about brands to advocate in favor of a taxonomy, I think it was a few weeks ago ) of the number of prices per "supermarket chain" ? i.e Carrefour City, Carrefour Contact, U Express, Proxi, Spar, Franprix, etc...to see if there's a way to optionally filter based on the size of the shop or something to that effect, prices from those small supermarkets create spikes on the graphs.

edit : was it what you meant by Location : type in the table up top ? I saw Location : type on the wiki.

[raphodn]

Here you go, I did a basic aggregation of price location's "osm_name"
the top 30 results are here : https://github.com/openfoodfacts/open-prices/wiki/Stats#top-location-names

[raphodn]

the shop names (at least their homogenization) happens mostly here : https://github.com/osmlab/name-suggestion-index

that's why i opened a thread in the OSM forum a couple of weeks ago ^^
https://forum.openstreetmap.fr/t/aidez-moi-a-lister-rajouter-des-chaines-de-magasins-manquants-dans-nsi/26727

[raphodn]

location type : "shop:supermarket"
location name : "Carrefour City" (added in the table 👍 )

[raphodn]

makes me think of how to make these stats easily accessible/queryable 🤔
(without having to run SQL or django shell commands)
dedicated endpoints ?

[serpico]

Packaging size : ex: using both https://prices.openfoodfacts.org/app/categories/en:cocoa-and-hazelnuts-spreads and https://prices.openfoodfacts.org/app/brands/Nutella to show price/kg evolution depending on packaging size another rather striking difference ( I wasn't expecting ) between 3760335580666 and 3760197946945.

Maybe adding brand to the list of filters under the table would be enough ?

[serpico]

Filtering through quantity : if I'm looking for Tea in bulk inside https://prices.openfoodfacts.org/app/categories/en:green-teas I have a hunch anything < 100 g won't return what I'm looking for .