A guide to the InfluxDBMapper and QueryBuilder for Java: Group By

Previously we executed some selection examples and aggregations against an InfluxDB database. In this tutorial we are going to check the group by functionality that the Query Builder provides to us with.

Before you start you need to spin up an influxdb instance with the data needed.

Supposing that we want to group by a single tag, we shall use the groupBy function.

        Query query = select().mean("water_level").from(DATABASE, "h2o_feet").groupBy("location");
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);

The query to be executed shall be

SELECT MEAN(water_level) FROM h2o_feet GROUP BY location;

If we want to group by multiple tags we will pass an array of tags.

        Query query = select().mean("index").from(DATABASE,"h2o_feet")
                              .groupBy("location","randtag");
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);

The result will be

SELECT MEAN(index) FROM h2o_feet GROUP BY location,randtag;

Another option is to query by all tags.

        Query query = select().mean("index").from(DATABASE,"h2o_feet")
                              .groupBy(raw("*"));
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);
SELECT MEAN(index) FROM h2o_feet GROUP BY *;

Since InfluxDB is a time series database we have great group by functionality based on time.

For example let’s group query results into 12 minute intervals

        Query query = select().count("water_level").from(DATABASE,"h2o_feet")
                              .where(eq("location","coyote_creek"))
                              .and(gte("time","2015-08-18T00:00:00Z"))
                              .and(lte("time","2015-08-18T00:30:00Z"))
                              .groupBy(time(12l,MINUTE));
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);

We get the result

SELECT COUNT(water_level) FROM h2o_feet WHERE location = 'coyote_creek' AND time >= '2015-08-18T00:00:00Z' AND time <= '2015-08-18T00:30:00Z' GROUP BY time(12m);

Group results by 12 minute intervals and location.

        Query query = select().count("water_level").from(DATABASE,"h2o_feet")
                              .where()
                              .and(gte("time","2015-08-18T00:00:00Z"))
                              .and(lte("time","2015-08-18T00:30:00Z"))
                              .groupBy(time(12l,MINUTE),"location");
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);

We get the following query.

SELECT COUNT(water_level) FROM h2o_feet WHERE time >= '2015-08-18T00:00:00Z' AND time <= '2015-08-18T00:30:00Z' GROUP BY time(12m),location;

We will get more advanced and group query results into 18 minute intervals and shift the preset time boundaries forward.

        Query query = select().mean("water_level").from(DATABASE,"h2o_feet")
                              .where(eq("location","coyote_creek"))
                              .and(gte("time","2015-08-18T00:06:00Z"))
                              .and(lte("time","2015-08-18T00:54:00Z"))
                              .groupBy(time(18l,MINUTE,6l,MINUTE));
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);
SELECT MEAN(water_level) FROM h2o_feet WHERE location = 'coyote_creek' AND time >= '2015-08-18T00:06:00Z' AND time <= '2015-08-18T00:54:00Z' GROUP BY time(18m,6m);

Or group query results into 12 minute intervals and shift the preset time boundaries back;

        Query query = select().mean("water_level").from(DATABASE,"h2o_feet")
                              .where(eq("location","coyote_creek"))
                              .and(gte("time","2015-08-18T00:06:00Z"))
                              .and(lte("time","2015-08-18T00:54:00Z"))
                              .groupBy(time(18l,MINUTE,-12l,MINUTE));
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);

The result would be

SELECT MEAN(water_level) FROM h2o_feet WHERE location = 'coyote_creek' AND time >= '2015-08-18T00:06:00Z' AND time <= '2015-08-18T00:54:00Z' GROUP BY time(18m,-12m);

Eventually we can group by and fill

        Query query = select()
                .column("water_level")
                .from(DATABASE, "h2o_feet")
                .where(gt("time", op(ti(24043524l, MINUTE), SUB, ti(6l, MINUTE))))
                .groupBy("water_level")
                .fill(100);
        LOGGER.info("Executing query "+query.getCommand());
        QueryResult queryResult = influxDB.query(query);

The result would be

SELECT water_level FROM h2o_feet WHERE time > 24043524m - 6m GROUP BY water_level fill(100);

That’s it! We just run some really complex group by queries against our InfluxDB database. The query builder makes it possible to create queries using only java.
You can find the sourcecode in github.