[ML] Explain Log Rate Spikes: Group results API. #140683
Conversation
walterra
added
:ml
release_note:skip
|
Pinging @elastic/ml-ui (:ml) |
walterra
changed the title
💚 Build Succeeded
Metrics [docs]Public APIs missing comments
Async chunks
History
To update your PR or re-run it, just comment with: cc @walterra |
| // get unique fields that are left | ||
| const fields = [...new Set(terms.map((t) => t.fieldName))]; | ||
|
|
||
| // TODO add query params |
There was a problem hiding this comment.
Will this be added in a follow up?
| */ | ||
|
|
||
| import { max } from 'd3-array'; | ||
| // import { omit, uniq } from 'lodash'; |
There was a problem hiding this comment.
Can remove this commented line
There was a problem hiding this comment.
I'll leave it in because it's part of another commented part around categories which we might need later.
|
|
||
| // TODO - add handling of creating * as next level of tree | ||
|
|
||
| if (Object.keys(getValueCounts(filteredItemSets, nextField)).length > 0) { |
There was a problem hiding this comment.
Any chance the break condition is never hit and we stay in an infinite while loop? Like could that length always be zero for some reason?
There was a problem hiding this comment.
This is approach is based on the original Python integration. I'll stick to it for now, once we have better test coverage we can maybe refactor to get rid of the while(true).
| displayOther: boolean, | ||
| fields: string[] = [] | ||
| ) { | ||
| // const candidates = uniq( |
There was a problem hiding this comment.
Nit: commented code
| leaves: ChangePointGroup[], | ||
| level = 1 | ||
| ) { | ||
| // console.log(`${'-'.repeat(level)} ${tree.name} ${tree.children.length}`); |
There was a problem hiding this comment.
Nit: can remove this comment
Summary
Part of #138117.
Extends the
/internal/aiops/explain_log_rate_spikeswith an option to extend the analysis and summarize significant field/value pairs into groups using thefrequent_itemsaggregation.Checklist