Collect ideas for statistical values of primary data

[geofranzi]

Idea: Calculate statistical values based on materialized views, after refresh with data. Store results in a new table in BEXIS as JSON. Create one DB function e.g. update_statistics(id) to run it.

dubclicates

select Column1, Column2, count(*) from table_name group by Column1, Column2 HAVING count(*) > 1 (use all var columns)

unique values + count

select Column1, count(id) from table_name group by Column1 (count by id to get correct count of empty/NULL fields)

all except of bool, date, string, ...(excluding placeholder values!)

• sum
• avg
• median
• min/max
• ...

total count

---

nice summary statistics for any data table: http://madlib.incubator.apache.org/docs/latest/group__grp__summary.html

[geofranzi]

• Add API Call: /api/DataStatistic/{id}/{VariableId} @DavidBlaa for Pawanand and others)

unique value:
select [VariableId], count(id) from mvdataset**[id]** group by var[VariableId]

count:
select count(*) from mvdataset[id]