# Non-aggregate workarounds

**URL:** <https://discuss.elastic.co/t/non-aggregate-workarounds/138815>\
**Category:** Kibana\
**Created:** [July 6, 2018, 2:40am UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815 "2018-07-06T02:40:54Z")\
**Posts on this page:** 8\
**Page:** 1

<div class="post-metadata">

**Author:** ![nonaggregates-thanks](https://avatars.discourse-cdn.com/v4/letter/n/46a35a/32.png) [@nonaggregates-thanks](https://discuss.elastic.co/u/nonaggregates-thanks)\
**Post date:** [July 6, 2018, 2:40am UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/1 "2018-07-06T02:40:54Z")

</div>

so I've been having some pretty massive difficulty with this. Here is the json format of what im visualizing

```
"properties": {
    "id" { "type" : "text" }, 
    "nested_map" : { "type" : "nested"}  
}

```

where the nested object contains the name of the process and the double of its rate, i.e.

```
{ "name" : {
    "type": "text" },
"rate": {
  "type" : "double"
  }
} 

```

I want to be able to display, per process name, the associated rate. x-axis is term, y-axis is the actual value corresponding to that term. It looks like aggregation is the only visualization that is supported. This sucks. Am I able to work around this somehow?

---

<div class="post-metadata">

**Author:** ![bhavyarm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/bhavyarm/32/22392_2.png) [@bhavyarm](https://discuss.elastic.co/u/bhavyarm)\
**Post date:** [July 6, 2018, 6:34pm UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/2 "2018-07-06T18:34:59Z")

</div>

Hi,

Can you please post a couple of sample documents and your mapping?  
I am pretty sure we do not support this yet. But would like to confirm.

Thanks,  
Bhavya

---

<div class="post-metadata">

**Author:** ![bhavyarm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/bhavyarm/32/22392_2.png) [@bhavyarm](https://discuss.elastic.co/u/bhavyarm)\
**Post date:** [July 6, 2018, 7:18pm UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/3 "2018-07-06T19:18:24Z")

</div>

Can you also please take a look at this discuss post? [Workaround for nested objects](https://discuss.elastic.co/t/workaround-for-nested-objects/107108/6)

Thanks,  
Bhavya

---

<div class="post-metadata">

**Author:** ![nonaggregates-thanks](https://avatars.discourse-cdn.com/v4/letter/n/46a35a/32.png) [@nonaggregates-thanks](https://discuss.elastic.co/u/nonaggregates-thanks)\
**Post date:** [July 10, 2018, 8:18pm UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/4 "2018-07-10T20:18:50Z")

</div>

I have tried many mappings in order to work around this, but none worked. default mappings are  
{EID : [{ "sId" : { "type" : "text" }, "process" : { "type" : "text"} }] }

That said, I did create an alternative version that was suggested through the explicit mapping page,  
{EID : { "type" : "nested" } }. Both didn't work

---

<div class="post-metadata">

**Author:** ![nonaggregates-thanks](https://avatars.discourse-cdn.com/v4/letter/n/46a35a/32.png) [@nonaggregates-thanks](https://discuss.elastic.co/u/nonaggregates-thanks)\
**Post date:** [July 10, 2018, 8:21pm UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/5 "2018-07-10T20:21:36Z")

</div>

I think kibana would be very helpful for lay-users if it supported a form of optional non-aggregate visualization, because then dashboards can be created on the fly on a per usecase demand

---

<div class="post-metadata">

**Author:** ![bhavyarm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/bhavyarm/32/22392_2.png) [@bhavyarm](https://discuss.elastic.co/u/bhavyarm)\
**Post date:** [July 10, 2018, 8:39pm UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/6 "2018-07-10T20:39:44Z")

</div>

Hi,

Yep. We have an existing bug: [https://github.com/elastic/kibana/issues/1084](https://github.com/elastic/kibana/issues/1084). Can you please subscribe to it for updates and comment with your usecase .

Thanks,  
Bhavya

---

<div class="post-metadata">

**Author:** ![nonaggregates-thanks](https://avatars.discourse-cdn.com/v4/letter/n/46a35a/32.png) [@nonaggregates-thanks](https://discuss.elastic.co/u/nonaggregates-thanks)\
**Post date:** [July 10, 2018, 8:43pm UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/7 "2018-07-10T20:43:15Z")

</div>

This is the lookup for the associated integers used in the sum buckets. {EID : [{ "process" : "text", "howfast" : "integer" }] } OR it could be {EID: { "processName" : "number" } } and then for each process name, it could have its own number. I think that is similar to the suggestion in the link you provided. I could hypothetically smash the two json together and then see if the objects sharing the "sId" parameter would be a way of associating them. I will give that a shot and let you know how it goes.

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [August 7, 2018, 8:43pm UTC](https://discuss.elastic.co/t/non-aggregate-workarounds/138815/8 "2018-08-07T20:43:17Z")

</div>

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.