BigQuery : Daily averages for time series data

Question

I have timeseries data for some routers in a network. I need to compute the daily throughput per category for data that looks like this. ( https://i.stack.imgur.com/PCvan.png )

Most of the examples online talk about moving averages over time. But I want the daily averages for every single day in the past year by router and network name. Something like

Expected output

I looked at AVG over a WINDOW. But it only gives the moving average for one fixed time period(1 day prior, 1 month prior etc). Is there a way for compute the averages for every single day in the year? Grouped by routerID and network name

Answer 1

Is this query what you need?

    SELECT
        DATE(ts) AS day,
        RouterID AS router_id,
        wifi_ssid,        
        AVERAGE(avg_kbps) AS average_bandwidth_for_day,
      FROM
        ts_table
  GROUP BY
        1, 2, 3

(if not, you will need to clarify your question)