Watcher with dynamic filter

[Jesus_Rodrigo (Jesús Rodrigo)]

Hi,

I'm trying to make a Watcher that compares if a number field of a document deviates from the average a percentage.

The mapping the following:

{
"Datapoint" : "Taller10#Montaje#Skillets#M111#RODILLOS_LONG#M111R",
"ACTUAL_CONSUMPTION" : 80
}

The problem is that I have to calculate the average of the ACTUAL_CONSUMPTION field of the logs that have the same value in the field "Datapoint" and not of the rest.

[spinscale (Alexander Reelsen)]

so you want to group by the value of the DataPoint field and then calculate the avg of the ACTUAL_CONSUMPTION field?

If you, you should check out aggregations, if you need to this for more than one value of the DataPoint field. You could use a terms aggregation and inside of that a avg agg.

If you only need the avg value of a single value of DataPoint then a filter query for that value and an avg aggregation are sufficient.

[Jesus_Rodrigo (Jesús Rodrigo)]

Hi,

First thank the help, I got the average DataPoint. My problem now is that I do not know how to compare the average with the ACTUAL_CONSUMPTION.

This is my watcher code:

{
"trigger": {
"schedule": {
"interval": "1d"
}
},
"input": {
"search": {
"request": {
"indices": "consum*",
"body": {
"query": {
"bool": {
"filter": {
"range": {
"ACTUAL_CONSUMPTIONst": {
"from": "now-5d",
"to": "now"
}
}
}
}
},
"aggs":{
"datapoint":{
"terms": {
"field": "Datapoint.keyword"
},
"aggs":{
"dp_avg": {
"avg":{
"field": "ACTUAL_CONSUMPTION",
"script": "_value1.1"
}
},
"dp_actual":{
"terms": {
"field": "ACTUAL_CONSUMPTION"
}
}
}
}
}
}
}
}
},
"condition":{
"array_compare":{
"ctx.payload.aggregations.datapoint.buckets.dp_actual.buckets":{
"path": "key",
"gte":{
"value": "{{ctx.payload.aggregations.datapoint.buckets.dp_avg}}"
}
}
}
},
"actions": {
"log": {
"logging": {
"level": "info",
"text": "*** Watcher para informar si un motor tiene una media anómala ****\n"
}
}
}
}

One part of the response is the following:

"aggregations" : {
"datapoint" : {
"doc_count_error_upper_bound" : 0,
"sum_other_doc_count" : 0,
"buckets" : [
{
"doc_count" : 4,
"dp_avg" : {
"value" : 52.25
},
"key" : "Taller10#Montaje#Skillets#M11#RODILLOS_LONG#M11",
"dp_actual" : {
"doc_count_error_upper_bound" : 0,
"sum_other_doc_count" : 0,
"buckets" : [
{
"doc_count" : 2,
"key" : 40
},
{
"doc_count" : 1,
"key" : 50
},
{
"doc_count" : 1,
"key" : 60
}
]
}
}

And I want know that Datapoint has a document with ACTUAL_CONSUMPTION > Average*1.1

Is possible do it??

Thank you!

[spinscale (Alexander Reelsen)]

hey,

you cannot use a compare condition for this to properly access the fields, you need to use a script condition.

--Alex

[Jesus_Rodrigo (Jesús Rodrigo)]

Thanks you for your help!

I resolved my problem using a script that compare two buckets and return a list with all Datapoint's names.