Mysql json datatype count and groupby

Question

So since mysql 5.7 i am now able to use the json datatype which is nice when your working with frameworks like angularjs.

Now since im fairly new to this (and this datat

Answer 1

Firstly you are storing all the JSON in one row, if you want to query the data like this each resident should have it's own row of his/her own JSON:

*other_fields*, {"A": 1, "B": "Debra", "C": "Peters", "D": "dpeters0@accuweather.com", "E": "Female", "F": "Tsagaan-Ovoo"}
*other_fields*, {"A": 2, "B": "Marc", "C": "Rasmussen", "D": "Marc@test.dk", "E": "Male", "F": "Copenhagen"}

Secondly you don't need to count the extracted data, you just need to count the rows for each country extracted. This will do:

SELECT JSON_EXTRACT(data, "$.F") AS country,
       COUNT(*) AS num_residents

Thirdly, you don't want to GROUP BY data at all as this is presumably unique per resident. This will leave you with a group for each row. I think you just want:

GROUP BY country

Putting it all together:

  SELECT JSON_EXTRACT(data, "$.F") AS country,
         COUNT(*) AS num_residents
    FROM kpi_data
   WHERE schema_id = 39
GROUP BY country

For decent performance, you may consider placing an secondary index on the extracted country.. see DOCS