This Grafana plugin looks like:

Disclaimer

This plugin has been tested with Grafana 10.x but there's still work to do before submitting to Grafana. New versions may require changes to the query or configuration.

Usage

Install:

$ sudo su -
# cd /var/lib/grafana/plugins
# git clone https://github.com/spectraphilic/grafana-windrose.git

Allow loading this plugin, in /etc/grafana/grafana.ini:

[plugins]
allow_loading_unsigned_plugins = spectraphilic-windrose-panel

Restart Grafana service.

Add a panel to your dashboard with the Windrose visualization type. Set up the query to provide data with speed and direction fields. The speed can be in the unit of your choice (can be set in the settings), while the direction is always in degrees.

PostgreSQL

Example query for PostgreSQL:

SELECT
  time,
  field1 AS speed,
  field2 AS direction
FROM table
WHERE $__unixEpochFilter("time")
ORDER BY time

ClickHouse

Example for ClickHouse:

SELECT
  $timeSeries as time,
  field1 AS speed,
  field2 AS direction
FROM table
WHERE $timeFilter
ORDER BY time

Prometheus

This is valid as well for prometheus-compatible data sources such as VictoriaMetrics.

Set a query for speed:

• Metrics browser: meteo_ws_mean{location="$location"}
• Legend: speed

Set another query for direction:

• Metrics browser: meteo_wd_mean{location="$location"}
• Legend: direction

This information was provided by @newrushbolt ; for further details see #18 (comment)

Flux

Example for InfluxDB v2 filled from SignalK.

from(bucket: "bucket")
  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)
  |> filter(fn: (r) => r["_measurement"] == "environment.wind.speedTrue" or r["_measurement"] == "environment.wind.directionTrue")
  |> filter(fn: (r) => r["_field"] == "value")
  |> pivot(rowKey:["_time"], columnKey: ["_measurement"], valueColumn: "_value") // put dir and speed values into the same row based on timestamps
  |> filter(fn: (r) => exists r["environment.wind.directionTrue"] and exists r["environment.wind.speedTrue"]) // drop rows that don't have both values
  |> rename(columns: {"environment.wind.directionTrue": "directionRad", "environment.wind.speedTrue": "speedMps"})
  |> map(fn: (r) => ({ r with  direction: r.directionRad / 3.14 * 180.0 })) // convert to degrees from radians
  |> map(fn: (r) => ({ r with  speed: r.speedMps / 0.514 })) // convert to knots from m/s
  |> aggregateWindow(every: v.windowPeriod, fn: first, column: "direction", createEmpty: false) // sample the first row for each window (real aggregation would show fake values in gusty winds)

Development

Install:

$ npm run dev
$ docker-compose up