I am using Date Histogram to aggregation values over time. The problem is that I see that if that date has no data, it just omits it completely. What I need it to do is enter it as null. I try to use missing
and also min_doc_count
set to 0 but the result is still the same.
Partial requests:
"aggs": {
"winner_over_time": {
"terms": {
"field": "winner",
"size": 10,
"order": {
"_term": "asc"
}
},
"aggs": {
"count_over_time": {
"date_histogram": {
"field": "timestamp",
"interval": interval,
"min_doc_count": 0,
"missing": 0
}
}
}
},
Partial response:
"aggregations": {
"winner_over_time": {
"doc_count_error_upper_bound": 0,
"sum_other_doc_count": 0,
"buckets": [
{
"key": -3,
"doc_count": 1,
"count_over_time": {
"buckets": [
{
"key_as_string": "2018-01-29T05:00:00.000Z",
"key": 1517202000000,
"doc_count": 1
}
]
}
},
{
"key": -1,
"doc_count": 3,
"count_over_time": {
"buckets": [
{
"key_as_string": "2018-01-29T00:00:00.000Z",
"key": 1517184000000,
"doc_count": 1
},
{
"key_as_string": "2018-01-29T01:00:00.000Z",
"key": 1517187600000,
"doc_count": 1
},
{
"key_as_string": "2018-01-29T02:00:00.000Z",
"key": 1517191200000,
"doc_count": 0
},
{
"key_as_string": "2018-01-29T03:00:00.000Z",
"key": 1517194800000,
"doc_count": 0
},
{
"key_as_string": "2018-01-29T04:00:00.000Z",
"key": 1517198400000,
"doc_count": 0
},
{
"key_as_string": "2018-01-29T05:00:00.000Z",
"key": 1517202000000,
"doc_count": 1
}
]
}
},
As you see, for "key": -3
, it only returned one data point for me. I need it to return everything within the series. Right now I had to go into the data and manually insert missing time keys for my charting library. I would prefer to not have to do that.