To get sensors values, use the command unit/calc_sensors:
svc=unit/calc_sensors¶ms={"source":<text>, "indexFrom":<uint>, "indexTo":<uint>, "unitId":<long>, "sensorId":<long>, "width":<uint>}
Name | Description |
---|---|
source | source of messages - layer name, if empty - messages loader will be used by default |
indexFrom | index of first message |
indexTo | index of last message |
unitId | unit ID |
sensorId | ID sensor: if 0 - return values of all sensors |
width | desired number of elements to respond, aggregation is used (see below), optional |
If source
is defined then unitId
must be defined too. If source
is NOT defined then write down any value in unitId
(it is not used) – messages are put from message loader directly.
When you need fixed quantity of sensor values and aggregated (i.e. preprocessed), then you may use 'width' param. For example you may get preprocessed data to form and zoom charts.
Logics is as follows:
1. the system finds all messages within asked bounds;
2. then it splits all the interval into width subintervals;
3. in every subinterval it finds the very first sensor value (left), the last one (right), minimal one (bottom) maximal one (top). Described values will merge into final server response.
[ /* array of messages data */ { /* value(s) of sensor(s) from one message */ <text>:<double|text>, /* sensor ID: sensor value */ ... } ]
If width is defined, then the response will be as follows:
[ [ { "name":<sensor_name>, "data":[ { "left":[<unix_time1>,<value1>], "right":[<unix_time2>,<value2>], "bottom":[<unix_time3>,<value3>], "top":[<unix_time4>,<value4>] }, ... ] }, ... ] ]
where <sensor_name> – sensor name,
<unix_timeN> – message UNIX-time,
valueN – sensor value in a message,
left – the first sensor value in a subinterval and its time,
right – the last sensor value in a subinterval and its time,
bottom – minimal sensor value in a subinterval and its time,
top – maximal sensor value in a subinterval and its time.
|