Help aggregating and writing to new table

Fluxlang
, , ,
4

Update:
I now have found a solution by experimenting with the experimental.to function.
Having changed the query to:

last = from(bucket: "testtelegraf")
    |> range(start: -7d)
    |> filter(fn: (r) => r["topic"] == "Import")
    |> aggregateWindow(every: 1d, fn: last, createEmpty: false)
first = from(bucket: "testtelegraf")
    |> range(start: -7d)
    |> filter(fn: (r) => r["topic"] == "Import")
    |> aggregateWindow(every: 1d, fn: first, createEmpty: false)

join(
    tables: {lastusage: last, firstusage: first},
    on: [
        "_time",
        "topic",
    ],
)
    |> map(
        fn: (r) => ({
            _value: r._value_lastusage - r._value_firstusage,
            _endvalue: r._value_lastusage,
            _time: r._time,
            _measurement: "solar_daily",
        }),
    )
    |> to(
  bucket:"test_daily",
  org:"my-org",
  fieldFn: (r) => ({"_value": r._value, "_endvalue": r._endvalue})
)

Now results in basically 2 tables on the test_daily bucket:

test_daily
test_daily1581×277 61.1 KB

Apparently the _field is now tagged with “_value” or “_endvalue” according to which value there is in the “_value” field.

So,I now have saved both the daily usage and the end-value in tables, and can use that in reports, so that is fine.

Still puzzled why I can’t see to create a single table with both values in each row, just like the result of the initial query. If somebody would show me how to do that (and how I seem not to understand to work with fields and tags), that would be great!