query language inspired by [PromQL](https://prometheus.io/docs/prometheus/latest/querying/basics/).
MetricsQL is backwards-compatible with PromQL, so Grafana dashboards backed by Prometheus datasource should work
the same after switching from Prometheus to VictoriaMetrics.
However, there are some [intentional differences](https://medium.com/@romanhavronenko/victoriametrics-promql-compliance-d4318203f51e) between these two languages.
[Standalone MetricsQL package](https://godoc.org/github.com/VictoriaMetrics/metricsql) can be used for parsing MetricsQL in external apps.
If you are unfamiliar with PromQL, then it is suggested reading [this tutorial for beginners](https://medium.com/@valyala/promql-tutorial-for-beginners-9ab455142085)
and introduction into [basic querying via MetricsQL](https://docs.victoriametrics.com/keyConcepts.html#metricsql).
* MetricsQL takes into account the last [raw sample](https://docs.victoriametrics.com/keyconcepts/#raw-samples) before the lookbehind window
in square brackets for [increase](#increase) and [rate](#rate) functions. This allows returning the exact results users expect for `increase(metric[$__interval])` queries
instead of incomplete results Prometheus returns for such queries. Prometheus misses the increase between the last sample before the lookbehind window
and the first sample inside the lookbehind window.
* MetricsQL doesn't extrapolate [rate](#rate) and [increase](#increase) function results, so it always returns the expected results. For example, it returns
integer results from `increase()` over slow-changing integer counter. Prometheus in this case returns unexpected fractional results,
which may significantly differ from the expected results. This addresses [this issue from Prometheus](https://github.com/prometheus/prometheus/issues/3746).
This addresses [this issue from Grafana](https://github.com/grafana/grafana/issues/11451).
See also [this blog post](https://www.percona.com/blog/2020/02/28/better-prometheus-rate-function-with-victoriametrics/).
* MetricsQL treats `scalar` type the same as `instant vector` without labels, since subtle differences between these types usually confuse users.
See [the corresponding Prometheus docs](https://prometheus.io/docs/prometheus/latest/querying/basics/#expression-language-data-types) for details.
* MetricsQL removes all the `NaN` values from the output, so some queries like `(-1)^0.5` return empty results in VictoriaMetrics,
while returning a series of `NaN` values in Prometheus. Note that Grafana doesn't draw any lines or dots for `NaN` values,
so the end result looks the same for both VictoriaMetrics and Prometheus.
* MetricsQL keeps metric names after applying functions, which don't change the meaning of the original time series.
For example, [min_over_time(foo)](#min_over_time) or [round(foo)](#round) leaves `foo` metric name in the result.
See [this issue](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/674) for details.
Read more about the differences between PromQL and MetricsQL in [this article](https://medium.com/@romanhavronenko/victoriametrics-promql-compliance-d4318203f51e).
Other PromQL functionality should work the same in MetricsQL.
[File an issue](https://github.com/VictoriaMetrics/VictoriaMetrics/issues) if you notice discrepancies between PromQL and MetricsQL results other than mentioned above.
and provides additional functionality mentioned below, which is aimed towards solving practical cases.
Feel free [filing a feature request](https://github.com/VictoriaMetrics/VictoriaMetrics/issues) if you think MetricsQL misses certain useful functionality.
This functionality can be evaluated at [VictoriaMetrics playground](https://play.victoriametrics.com/select/accounting/1/6a716b0f-38bc-4856-90ce-448fd713e3fe/prometheus/graph/)
or at your own [VictoriaMetrics instance](https://docs.victoriametrics.com/#how-to-start-victoriametrics).
The list of MetricsQL features on top of PromQL:
* Graphite-compatible filters can be passed via `{__graphite__="foo.*.bar"}` syntax.
See [these docs](https://docs.victoriametrics.com/#selecting-graphite-metrics).
* Lookbehind window in square brackets for [rollup functions](#rollup-functions) may be omitted. VictoriaMetrics automatically selects the lookbehind window
depending on the `step` query arg passed to [/api/v1/query_range](https://docs.victoriametrics.com/keyConcepts.html#range-query)
and the real interval between [raw samples](https://docs.victoriametrics.com/keyconcepts/#raw-samples) (aka `scrape_interval`).
* Numeric values can contain `_` delimiters for better readability. For example, `1_234_567_890` can be used in queries instead of `1234567890`.
* [Series selectors](https://docs.victoriametrics.com/keyConcepts.html#filtering) accept multiple `or` filters. For example, `{env="prod",job="a" or env="dev",job="b"}`
See [these docs](https://docs.victoriametrics.com/keyConcepts.html#filtering-by-multiple-or-filters) for details.
* Support for `group_left(*)` and `group_right(*)` for copying all the labels from time series on the `one` side
of [many-to-one operations](https://prometheus.io/docs/prometheus/latest/querying/operators/#many-to-one-and-one-to-many-vector-matches).
The copied label names may clash with the existing label names, so MetricsQL provides an ability to add prefix to the copied metric names
via `group_left(*) prefix "..."` syntax.
For example, the following query copies all the `namespace`-related labels from `kube_namespace_labels` to `kube_pod_info` series,
while adding `ns_` prefix to the copied labels: `kube_pod_info * on(namespace) group_left(*) prefix "ns_" kube_namespace_labels`.
Labels from the `on()` list aren't copied.
* [Aggregate functions](#aggregate-functions) accept arbitrary number of args.
For example, `avg(q1, q2, q3)` would return the average values for every point across time series returned by `q1`, `q2` and `q3`.
* [@ modifier](https://prometheus.io/docs/prometheus/latest/querying/basics/#modifier) can be put anywhere in the query.
For example, `sum(foo) @ end()` calculates `sum(foo)` at the `end` timestamp of the selected time range `[start ... end]`.
* Arbitrary subexpression can be used as [@ modifier](https://prometheus.io/docs/prometheus/latest/querying/basics/#modifier).
For example, `foo @ (end() - 1h)` calculates `foo` at the `end - 1 hour` timestamp on the selected time range `[start ... end]`.
* [offset](https://prometheus.io/docs/prometheus/latest/querying/basics/#offset-modifier), lookbehind window in square brackets
and `step` value for [subquery](#subqueries) may refer to the current step aka `$__interval` value from Grafana with `[Ni]` syntax.
For instance, `rate(metric[10i] offset 5i)` would return per-second rate over a range covering 10 previous steps with the offset of 5 steps.
* [offset](https://prometheus.io/docs/prometheus/latest/querying/basics/#offset-modifier) may be put anywhere in the query. For instance, `sum(foo) offset 24h`.
* Lookbehind window in square brackets and [offset](https://prometheus.io/docs/prometheus/latest/querying/basics/#offset-modifier) may be fractional.
For instance, `rate(node_network_receive_bytes_total[1.5m] offset 0.5d)`.
* The duration suffix is optional. The duration is in seconds if the suffix is missing.
For example, `rate(m[300] offset 1800)` is equivalent to `rate(m[5m]) offset 30m`.
* The duration can be placed anywhere in the query. For example, `sum_over_time(m[1h]) / 1h` is equivalent to `sum_over_time(m[1h]) / 3600`.
* Numeric values can have `K`, `Ki`, `M`, `Mi`, `G`, `Gi`, `T` and `Ti` suffixes. For example, `8K` is equivalent to `8000`, while `1.2Mi` is equivalent to `1.2*1024*1024`.
* Trailing commas on all the lists are allowed - label filters, function args and with expressions.
For instance, the following queries are valid: `m{foo="bar",}`, `f(a, b,)`, `WITH (x=y,) x`.
This simplifies maintenance of multi-line queries.
* Metric names and label names may contain any unicode letter. For example `температура{город="Киев"}` is a value MetricsQL expression.
* Metric names and labels names may contain escaped chars. For example, `foo\-bar{baz\=aa="b"}` is valid expression.
It returns time series with name `foo-bar` containing label `baz=aa` with value `b`.
Additionally, the following escape sequences are supported:
-`\xXX`, where `XX` is hexadecimal representation of the escaped ascii char.
-`\uXXXX`, where `XXXX` is a hexadecimal representation of the escaped unicode char.
* Aggregate functions support optional `limit N` suffix in order to limit the number of output series.
For example, `sum(x) by (y) limit 3` limits the number of output time series after the aggregation to 3.
All the other time series are dropped.
* [histogram_quantile](#histogram_quantile) accepts optional third arg - `boundsLabel`.
In this case it returns `lower` and `upper` bounds for the estimated percentile.
See [this issue for details](https://github.com/prometheus/prometheus/issues/5706).
*`default` binary operator. `q1 default q2` fills gaps in `q1` with the corresponding values from `q2`. See also [drop_empty_series](#drop_empty_series).
*`if` binary operator. `q1 if q2` removes values from `q1` for missing values from `q2`.
*`ifnot` binary operator. `q1 ifnot q2` removes values from `q1` for existing values from `q2`.
*`WITH` templates. This feature simplifies writing and managing complex queries.
Go to [WITH templates playground](https://play.victoriametrics.com/select/accounting/1/6a716b0f-38bc-4856-90ce-448fd713e3fe/expand-with-exprs) and try it.
* String literals may be concatenated. This is useful with `WITH` templates:
This modifier prevents from dropping metric names in function results. See [these docs](#keep_metric_names).
## keep_metric_names
By default, metric names are dropped after applying functions or [binary operators](https://prometheus.io/docs/prometheus/latest/querying/operators/#binary-operators),
since they may change the meaning of the original time series.
This may result in `duplicate time series` error when the function is applied to multiple time series with different names.
This error can be fixed by applying `keep_metric_names` modifier to the function or binary operator.
For example:
-`rate({__name__=~"foo|bar"}) keep_metric_names` leaves `foo` and `bar` metric names in the returned time series.
-`({__name__=~"foo|bar"} / 10) keep_metric_names` leaves `foo` and `bar` metric names in the returned time series.
## MetricsQL functions
If you are unfamiliar with PromQL, then please read [this tutorial](https://medium.com/@valyala/promql-tutorial-for-beginners-9ab455142085) at first.
**Rollup functions** (aka range functions or window functions) calculate rollups over **raw samples**
on the given lookbehind window for the [selected time series](https://docs.victoriametrics.com/keyConcepts.html#filtering).
For example, `avg_over_time(temperature[24h])` calculates the average temperature over raw samples for the last 24 hours.
Additional details:
* If rollup functions are used for building graphs in Grafana, then the rollup is calculated independently per each point on the graph.
For example, every point for `avg_over_time(temperature[24h])` graph shows the average temperature for the last 24 hours ending at this point.
The interval between points is set as `step` query arg passed by Grafana to [/api/v1/query_range](https://docs.victoriametrics.com/keyConcepts.html#range-query).
* If the given [series selector](https://docs.victoriametrics.com/keyConcepts.html#filtering) returns multiple time series,
then rollups are calculated individually per each returned series.
* If lookbehind window in square brackets is missing, then it is automatically set to the following value:
- To `step` value passed to [/api/v1/query_range](https://docs.victoriametrics.com/keyConcepts.html#range-query) or [/api/v1/query](https://docs.victoriametrics.com/keyconcepts/#instant-query)
for all the [rollup functions](#rollup-functions) except of [default_rollup](#default_rollup) and [rate](#rate). This value is known as `$__interval` in Grafana or `1i` in MetricsQL.
For example, `avg_over_time(temperature)` is automatically transformed to `avg_over_time(temperature[1i])`.
- To the `max(step, scrape_interval)`, where `scrape_interval` is the interval between [raw samples](https://docs.victoriametrics.com/keyconcepts/#raw-samples)
for [default_rollup](#default_rollup) and [rate](#rate) functions. This allows avoiding unexpected gaps on the graph when `step` is smaller than `scrape_interval`.
`count_ne_over_time(series_selector[d], ne)` is a [rollup function](#rollup-functions), which calculates the number of raw samples
on the given lookbehind window `d`, which aren't equal to `ne`. It is calculated independently per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [count_over_time](#count_over_time).
#### count_over_time
`count_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the number of raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [count_le_over_time](#count_le_over_time), [count_gt_over_time](#count_gt_over_time), [count_eq_over_time](#count_eq_over_time) and [count_ne_over_time](#count_ne_over_time).
#### count_values_over_time
`count_values_over_time("label", series_selector[d])` is a [rollup function](#rollup-functions), which counts the number of raw samples
with the same value over the given lookbehind window and stores the counts in a time series with an additional `label`, which contains each initial value.
The results are calculated independently per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [count_eq_over_time](#count_eq_over_time), [count_values](#count_values) and [distinct_over_time](#distinct_over_time) and [label_match](#label_match).
`decreases_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the number of raw sample value decreases
over the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [increases_over_time](#increases_over_time).
#### default_rollup
`default_rollup(series_selector[d])` is a [rollup function](#rollup-functions), which returns the last raw sample value on the given lookbehind window `d`
per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
If the lookbehind window is skipped in square brackets, then it is automatically calculated as `max(step, scrape_interval)`, where `step` is the query arg value
passed to [/api/v1/query_range](https://docs.victoriametrics.com/keyconcepts/#range-query) or [/api/v1/query](https://docs.victoriametrics.com/keyconcepts/#instant-query),
while `scrape_interval` is the interval between [raw samples](https://docs.victoriametrics.com/keyconcepts/#raw-samples) for the selected time series.
This allows avoiding unexpected gaps on the graph when `step` is smaller than the `scrape_interval`.
`deriv_fast(series_selector[d])` is a [rollup function](#rollup-functions), which calculates per-second derivative
using the first and the last raw samples on the given lookbehind window `d` per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [deriv](#deriv) and [ideriv](#ideriv).
#### descent_over_time
`descent_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates descent of raw sample values
on the given lookbehind window `d`. The calculations are performed individually per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
This function is useful for tracking height loss in GPS tracking.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [ascent_over_time](#ascent_over_time).
#### distinct_over_time
`distinct_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the number of distinct raw sample values
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`duration_over_time(series_selector[d], max_interval)` is a [rollup function](#rollup-functions), which returns the duration in seconds
when time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering) were present
over the given lookbehind window `d`. It is expected that intervals between adjacent samples per each series don't exceed the `max_interval`.
Otherwise, such intervals are considered as gaps and aren't counted.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [lifetime](#lifetime) and [lag](#lag).
#### first_over_time
`first_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the first raw sample value
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
See also [last_over_time](#last_over_time) and [tfirst_over_time](#tfirst_over_time).
#### geomean_over_time
`geomean_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates [geometric mean](https://en.wikipedia.org/wiki/Geometric_mean)
over raw samples on the given lookbehind window `d` per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### histogram_over_time
`histogram_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates
[VictoriaMetrics histogram](https://godoc.org/github.com/VictoriaMetrics/metrics#Histogram) over raw samples on the given lookbehind window `d`.
It is calculated individually per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
The resulting histograms are useful to pass to [histogram_quantile](#histogram_quantile) for calculating quantiles
over multiple [gauges](https://docs.victoriametrics.com/keyConcepts.html#gauge).
For example, the following query calculates median temperature by country over the last 24 hours:
`histogram_quantile(0.5, sum(histogram_over_time(temperature[24h])) by (vmrange,country))`.
#### hoeffding_bound_lower
`hoeffding_bound_lower(phi, series_selector[d])` is a [rollup function](#rollup-functions), which calculates
lower [Hoeffding bound](https://en.wikipedia.org/wiki/Hoeffding%27s_inequality) for the given `phi` in the range `[0...1]`.
See also [hoeffding_bound_upper](#hoeffding_bound_upper).
#### hoeffding_bound_upper
`hoeffding_bound_upper(phi, series_selector[d])` is a [rollup function](#rollup-functions), which calculates
upper [Hoeffding bound](https://en.wikipedia.org/wiki/Hoeffding%27s_inequality) for the given `phi` in the range `[0...1]`.
See also [hoeffding_bound_lower](#hoeffding_bound_lower).
#### holt_winters
`holt_winters(series_selector[d], sf, tf)` is a [rollup function](#rollup-functions), which calculates Holt-Winters value
(aka [double exponential smoothing](https://en.wikipedia.org/wiki/Exponential_smoothing#Double_exponential_smoothing)) for raw samples
over the given lookbehind window `d` using the given smoothing factor `sf` and the given trend factor `tf`.
Both `sf` and `tf` must be in the range `[0...1]`. It is expected that the [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering)
returns time series of [gauge type](https://docs.victoriametrics.com/keyConcepts.html#gauge).
`idelta(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the difference between the last two raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`increase_prometheus(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the increase
over the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
It is expected that the `series_selector` returns time series of [counter type](https://docs.victoriametrics.com/keyConcepts.html#counter).
It doesn't take into account the last sample before the given lookbehind window `d` when calculating the result in the same way as Prometheus does.
See [this article](https://medium.com/@romanhavronenko/victoriametrics-promql-compliance-d4318203f51e) for details.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [increase_pure](#increase_pure) and [increase](#increase).
#### increase_pure
`increase_pure(series_selector[d])` is a [rollup function](#rollup-functions), which works the same as [increase](#increase) except
of the following corner case - it assumes that [counters](https://docs.victoriametrics.com/keyConcepts.html#counter) always start from 0,
while [increase](#increase) ignores the first value in a series if it is too big.
#### increases_over_time
`increases_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the number of raw sample value increases
over the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [decreases_over_time](#decreases_over_time).
#### integrate
`integrate(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the integral over raw samples on the given lookbehind window `d`
per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### irate
`irate(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the "instant" per-second increase rate over the last two raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
It is expected that the `series_selector` returns time series of [counter type](https://docs.victoriametrics.com/keyConcepts.html#counter).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`lag(series_selector[d])` is a [rollup function](#rollup-functions), which returns the duration in seconds between the last sample
on the given lookbehind window `d` and the timestamp of the current point. It is calculated independently per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [lifetime](#lifetime) and [duration_over_time](#duration_over_time).
#### last_over_time
`last_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the last raw sample value on the given lookbehind window `d`
per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
`lifetime(series_selector[d])` is a [rollup function](#rollup-functions), which returns the duration in seconds between the last and the first sample
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [duration_over_time](#duration_over_time) and [lag](#lag).
#### mad_over_time
`mad_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates [median absolute deviation](https://en.wikipedia.org/wiki/Median_absolute_deviation)
over raw samples on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
`max_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the maximum value over raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
`median_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates median value over raw samples
on the given lookbehind window `d` per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
See also [avg_over_time](#avg_over_time).
#### min_over_time
`min_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the minimum value over raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
`mode_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates [mode](https://en.wikipedia.org/wiki/Mode_(statistics))
for raw samples on the given lookbehind window `d`. It is calculated individually per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering). It is expected that raw sample values are discrete.
`outlier_iqr_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the last sample on the given lookbehind window `d`
if its value is either smaller than the `q25-1.5*iqr` or bigger than `q75+1.5*iqr` where:
-`iqr` is an [Interquartile range](https://en.wikipedia.org/wiki/Interquartile_range) over raw samples on the lookbehind window `d`
-`q25` and `q75` are 25th and 75th [percentiles](https://en.wikipedia.org/wiki/Percentile) over raw samples on the lookbehind window `d`.
The `outlier_iqr_over_time()` is useful for detecting anomalies in gauge values based on the previous history of values.
For example, `outlier_iqr_over_time(memory_usage_bytes[1h])` triggers when `memory_usage_bytes` suddenly goes outside the usual value range for the last 24 hours.
`present_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns 1 if there is at least a single raw sample
on the given lookbehind window `d`. Otherwise, an empty result is returned.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
This function is supported by PromQL.
#### quantile_over_time
`quantile_over_time(phi, series_selector[d])` is a [rollup function](#rollup-functions), which calculates `phi`-quantile over raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
`quantiles_over_time("phiLabel", phi1, ..., phiN, series_selector[d])` is a [rollup function](#rollup-functions), which calculates `phi*`-quantiles
over raw samples on the given lookbehind window `d` per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
The function returns individual series per each `phi*` with `{phiLabel="phi*"}` label. `phi*` values must be in the range `[0...1]`.
See also [quantile_over_time](#quantile_over_time).
#### range_over_time
`range_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates value range over raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
E.g. it calculates `max_over_time(series_selector[d]) - min_over_time(series_selector[d])`.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### rate
`rate(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the average per-second increase rate
over the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
It is expected that the `series_selector` returns time series of [counter type](https://docs.victoriametrics.com/keyConcepts.html#counter).
If the lookbehind window is skipped in square brackets, then it is automatically calculated as `max(step, scrape_interval)`, where `step` is the query arg value
passed to [/api/v1/query_range](https://docs.victoriametrics.com/keyconcepts/#range-query) or [/api/v1/query](https://docs.victoriametrics.com/keyconcepts/#instant-query),
while `scrape_interval` is the interval between [raw samples](https://docs.victoriametrics.com/keyconcepts/#raw-samples) for the selected time series.
This allows avoiding unexpected gaps on the graph when `step` is smaller than the `scrape_interval`.
`rate_over_sum(series_selector[d])` is a [rollup function](#rollup-functions), which calculates per-second rate over the sum of raw samples
on the given lookbehind window `d`. The calculations are performed individually per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### resets
`resets(series_selector[d])` is a [rollup function](#rollup-functions), which returns the number
of [counter](https://docs.victoriametrics.com/keyConcepts.html#counter) resets over the given lookbehind window `d`
per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
It is expected that the `series_selector` returns time series of [counter type](https://docs.victoriametrics.com/keyConcepts.html#counter).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
This function is supported by PromQL.
#### rollup
`rollup(series_selector[d])` is a [rollup function](#rollup-functions), which calculates `min`, `max` and `avg` values for raw samples
on the given lookbehind window `d` and returns them in time series with `rollup="min"`, `rollup="max"` and `rollup="avg"` additional labels.
These values are calculated individually per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Optional 2nd argument `"min"`, `"max"` or `"avg"` can be passed to keep only one calculation result and without adding a label.
`rollup_candlestick(series_selector[d])` is a [rollup function](#rollup-functions), which calculates `open`, `high`, `low` and `close` values (aka OHLC)
over raw samples on the given lookbehind window `d` and returns them in time series with `rollup="open"`, `rollup="high"`, `rollup="low"` and `rollup="close"` additional labels.
The calculations are performed individually per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering). This function is useful for financial applications.
`rollup_delta(series_selector[d])` is a [rollup function](#rollup-functions), which calculates differences between adjacent raw samples
on the given lookbehind window `d` and returns `min`, `max` and `avg` values for the calculated differences
and returns them in time series with `rollup="min"`, `rollup="max"` and `rollup="avg"` additional labels.
The calculations are performed individually per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Optional 2nd argument `"min"`, `"max"` or `"avg"` can be passed to keep only one calculation result and without adding a label.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [rollup_increase](#rollup_increase).
#### rollup_deriv
`rollup_deriv(series_selector[d])` is a [rollup function](#rollup-functions), which calculates per-second derivatives
for adjacent raw samples on the given lookbehind window `d` and returns `min`, `max` and `avg` values for the calculated per-second derivatives
and returns them in time series with `rollup="min"`, `rollup="max"` and `rollup="avg"` additional labels.
The calculations are performed individually per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Optional 2nd argument `"min"`, `"max"` or `"avg"` can be passed to keep only one calculation result and without adding a label.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### rollup_increase
`rollup_increase(series_selector[d])` is a [rollup function](#rollup-functions), which calculates increases for adjacent raw samples
on the given lookbehind window `d` and returns `min`, `max` and `avg` values for the calculated increases
and returns them in time series with `rollup="min"`, `rollup="max"` and `rollup="avg"` additional labels.
The calculations are performed individually per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Optional 2nd argument `"min"`, `"max"` or `"avg"` can be passed to keep only one calculation result and without adding a label.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names. See also [rollup_delta](#rollup_delta).
#### rollup_rate
`rollup_rate(series_selector[d])` is a [rollup function](#rollup-functions), which calculates per-second change rates for adjacent raw samples
on the given lookbehind window `d` and returns `min`, `max` and `avg` values for the calculated per-second change rates
and returns them in time series with `rollup="min"`, `rollup="max"` and `rollup="avg"` additional labels.
See [this article](https://valyala.medium.com/why-irate-from-prometheus-doesnt-capture-spikes-45f9896d7832) in order to understand better
when to use `rollup_rate()`.
Optional 2nd argument `"min"`, `"max"` or `"avg"` can be passed to keep only one calculation result and without adding a label.
The calculations are performed individually per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### rollup_scrape_interval
`rollup_scrape_interval(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the interval in seconds between
adjacent raw samples on the given lookbehind window `d` and returns `min`, `max` and `avg` values for the calculated interval
and returns them in time series with `rollup="min"`, `rollup="max"` and `rollup="avg"` additional labels.
The calculations are performed individually per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Optional 2nd argument `"min"`, `"max"` or `"avg"` can be passed to keep only one calculation result and without adding a label.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names. See also [scrape_interval](#scrape_interval).
#### scrape_interval
`scrape_interval(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the average interval in seconds between raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [rollup_scrape_interval](#rollup_scrape_interval).
#### share_gt_over_time
`share_gt_over_time(series_selector[d], gt)` is a [rollup function](#rollup-functions), which returns share (in the range `[0...1]`) of raw samples
on the given lookbehind window `d`, which are bigger than `gt`. It is calculated independently per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
This function is useful for calculating SLI and SLO. Example: `share_gt_over_time(up[24h], 0)` - returns service availability for the last 24 hours.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`stale_samples_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the number
of [staleness markers](https://docs.victoriametrics.com/vmagent.html#prometheus-staleness-markers) on the given lookbehind window `d`
per each time series matching the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### stddev_over_time
`stddev_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates standard deviation over raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`stdvar_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates standard variance over raw samples
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`sum_eq_over_time(series_selector[d], eq)` is a [rollup function](#rollup-function), which calculates the sum of raw sample values equal to `eq`
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [sum_over_time](#sum_over_time) and [count_eq_over_time](#count_eq_over_time).
#### sum_gt_over_time
`sum_gt_over_time(series_selector[d], gt)` is a [rollup function](#rollup-function), which calculates the sum of raw sample values bigger than `gt`
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [sum_over_time](#sum_over_time) and [count_gt_over_time](#count_gt_over_time).
#### sum_le_over_time
`sum_le_over_time(series_selector[d], le)` is a [rollup function](#rollup-function), which calculates the sum of raw sample values smaller or equal to `le`
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [sum_over_time](#sum_over_time) and [count_le_over_time](#count_le_over_time).
`sum_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the sum of raw sample values
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
This function is supported by PromQL.
#### sum2_over_time
`sum2_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which calculates the sum of squares for raw sample values
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`timestamp(series_selector[d])` is a [rollup function](#rollup-functions), which returns the timestamp in seconds with millisecond precision for the last raw sample
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`timestamp_with_name(series_selector[d])` is a [rollup function](#rollup-functions), which returns the timestamp in seconds with millisecond precision for the last raw sample
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are preserved in the resulting rollups.
`tfirst_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the timestamp in seconds with millisecond precision for the first raw sample
on the given lookbehind window `d` per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`tlast_change_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the timestamp in seconds with millisecond precision for the last change
per each time series returned from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering) on the given lookbehind window `d`.
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [last_over_time](#last_over_time).
#### tlast_over_time
`tlast_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which is an alias for [timestamp](#timestamp).
See also [tlast_change_over_time](#tlast_change_over_time).
`tmax_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the timestamp in seconds with millisecond precision for the raw sample
`tmin_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns the timestamp in seconds with millisecond precision for the raw sample
with the minimum value on the given lookbehind window `d`. It is calculated independently per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
See also [min_over_time](#min_over_time).
#### zscore_over_time
`zscore_over_time(series_selector[d])` is a [rollup function](#rollup-functions), which returns [z-score](https://en.wikipedia.org/wiki/Standard_score)
for raw samples on the given lookbehind window `d`. It is calculated independently per each time series returned
from the given [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering).
Metric names are stripped from the resulting rollups. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`acos(q)` is a [transform function](#transform-functions), which returns [inverse cosine](https://en.wikipedia.org/wiki/Inverse_trigonometric_functions)
for every point of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`acosh(q)` is a [transform function](#transform-functions), which returns
[inverse hyperbolic cosine](https://en.wikipedia.org/wiki/Inverse_hyperbolic_functions#Inverse_hyperbolic_cosine) for every point of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`asinh(q)` is a [transform function](#transform-functions), which returns
[inverse hyperbolic sine](https://en.wikipedia.org/wiki/Inverse_hyperbolic_functions#Inverse_hyperbolic_sine) for every point of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`atan(q)` is a [transform function](#transform-functions), which returns [inverse tangent](https://en.wikipedia.org/wiki/Inverse_trigonometric_functions)
for every point of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`atanh(q)` is a [transform function](#transform-functions), which returns
[inverse hyperbolic tangent](https://en.wikipedia.org/wiki/Inverse_hyperbolic_functions#Inverse_hyperbolic_tangent) for every point of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`bitmap_and(q, mask)` is a [transform function](#transform-functions), which calculates bitwise `v & mask` for every `v` point of every time series returned from `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### bitmap_or
`bitmap_or(q, mask)` is a [transform function](#transform-functions), which calculates bitwise `v | mask` for every `v` point of every time series returned from `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### bitmap_xor
`bitmap_xor(q, mask)` is a [transform function](#transform-functions), which calculates bitwise `v ^ mask` for every `v` point of every time series returned from `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
#### buckets_limit
`buckets_limit(limit, buckets)` is a [transform function](#transform-functions), which limits the number
of [histogram buckets](https://valyala.medium.com/improving-histogram-usability-for-prometheus-and-grafana-bc7e5df0e350) to the given `limit`.
See also [prometheus_buckets](#prometheus_buckets) and [histogram_quantile](#histogram_quantile).
#### ceil
`ceil(q)` is a [transform function](#transform-functions), which rounds every point for every time series returned by `q` to the upper nearest integer.
`clamp(q, min, max)` is a [transform function](#transform-functions), which clamps every point for every time series returned by `q` with the given `min` and `max` values.
`clamp_max(q, max)` is a [transform function](#transform-functions), which clamps every point for every time series returned by `q` with the given `max` value.
`clamp_min(q, min)` is a [transform function](#transform-functions), which clamps every point for every time series returned by `q` with the given `min` value.
`day_of_month(q)` is a [transform function](#transform-functions), which returns the day of month for every point of every time series returned by `q`.
It is expected that `q` returns unix timestamps. The returned values are in the range `[1...31]`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`floor(q)` is a [transform function](#transform-functions), which rounds every point for every time series returned by `q` to the lower nearest integer.
`histogram_avg(buckets)` is a [transform function](#transform-functions), which calculates the average value for the given `buckets`.
It can be used for calculating the average over the given time range across multiple time series.
For example, `histogram_avg(sum(histogram_over_time(response_time_duration_seconds[5m])) by (vmrange,job))` would return the average response time
per each `job` over the last 5 minutes.
#### histogram_quantile
`histogram_quantile(phi, buckets)` is a [transform function](#transform-functions), which calculates `phi`-[percentile](https://en.wikipedia.org/wiki/Percentile)
over the given [histogram buckets](https://valyala.medium.com/improving-histogram-usability-for-prometheus-and-grafana-bc7e5df0e350).
`phi` must be in the range `[0...1]`. For example, `histogram_quantile(0.5, sum(rate(http_request_duration_seconds_bucket[5m])) by (le))`
would return median request duration for all the requests during the last 5 minutes.
The function accepts optional third arg - `boundsLabel`. In this case it returns `lower` and `upper` bounds for the estimated percentile with the given `boundsLabel` label.
See [this issue for details](https://github.com/prometheus/prometheus/issues/5706).
When the [percentile](https://en.wikipedia.org/wiki/Percentile) is calculated over multiple histograms,
then all the input histograms **must** have buckets with identical boundaries, e.g. they must have the same set of `le` or `vmrange` labels.
Otherwise, the returned result may be invalid. See [this issue](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3231) for details.
`histogram_quantiles("phiLabel", phi1, ..., phiN, buckets)` is a [transform function](#transform-functions), which calculates the given `phi*`-quantiles
over the given [histogram buckets](https://valyala.medium.com/improving-histogram-usability-for-prometheus-and-grafana-bc7e5df0e350).
Argument `phi*` must be in the range `[0...1]`. For example, `histogram_quantiles('le', 0.3, 0.5, sum(rate(http_request_duration_seconds_bucket[5m]) by (le))`.
Each calculated quantile is returned in a separate time series with the corresponding `{phiLabel="phi*"}` label.
See also [histogram_quantile](#histogram_quantile).
#### histogram_share
`histogram_share(le, buckets)` is a [transform function](#transform-functions), which calculates the share (in the range `[0...1]`)
for `buckets` that fall below `le`. This function is useful for calculating SLI and SLO. This is inverse to [histogram_quantile](#histogram_quantile).
The function accepts optional third arg - `boundsLabel`. In this case it returns `lower` and `upper` bounds for the estimated share with the given `boundsLabel` label.
#### histogram_stddev
`histogram_stddev(buckets)` is a [transform function](#transform-functions), which calculates standard deviation for the given `buckets`.
#### histogram_stdvar
`histogram_stdvar(buckets)` is a [transform function](#transform-functions), which calculates standard variance for the given `buckets`.
It can be used for calculating standard deviation over the given time range across multiple time series.
For example, `histogram_stdvar(sum(histogram_over_time(temperature[24])) by (vmrange,country))` would return standard deviation
for the temperature per each country over the last 24 hours.
#### hour
`hour(q)` is a [transform function](#transform-functions), which returns the hour for every point of every time series returned by `q`.
It is expected that `q` returns unix timestamps. The returned values are in the range `[0...23]`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
This function is supported by PromQL.
#### interpolate
`interpolate(q)` is a [transform function](#transform-functions), which fills gaps with linearly interpolated values calculated
from the last and the next non-empty points per each time series returned by `q`.
See also [keep_last_value](#keep_last_value) and [keep_next_value](#keep_next_value).
#### keep_last_value
`keep_last_value(q)` is a [transform function](#transform-functions), which fills gaps with the value of the last non-empty point
in every time series returned by `q`.
See also [keep_next_value](#keep_next_value) and [interpolate](#interpolate).
#### keep_next_value
`keep_next_value(q)` is a [transform function](#transform-functions), which fills gaps with the value of the next non-empty point
in every time series returned by `q`.
See also [keep_last_value](#keep_last_value) and [interpolate](#interpolate).
#### limit_offset
`limit_offset(limit, offset, q)` is a [transform function](#transform-functions), which skips `offset` time series from series returned by `q`
and then returns up to `limit` of the remaining time series per each group.
This allows implementing simple paging for `q` time series. See also [limitk](#limitk).
#### ln
`ln(q)` is a [transform function](#transform-functions), which calculates `ln(v)` for every point `v` of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`prometheus_buckets(buckets)` is a [transform function](#transform-functions), which converts
[VictoriaMetrics histogram buckets](https://valyala.medium.com/improving-histogram-usability-for-prometheus-and-grafana-bc7e5df0e350) with `vmrange` labels
to Prometheus histogram buckets with `le` labels. This may be useful for building heatmaps in Grafana.
See also [histogram_quantile](#histogram_quantile) and [buckets_limit](#buckets_limit).
#### rand
`rand(seed)` is a [transform function](#transform-functions), which returns pseudo-random numbers on the range `[0...1]` with even distribution.
Optional `seed` can be used as a seed for pseudo-random number generator.
See also [rand_normal](#rand_normal) and [rand_exponential](#rand_exponential).
#### rand_exponential
`rand_exponential(seed)` is a [transform function](#transform-functions), which returns pseudo-random numbers
with [exponential distribution](https://en.wikipedia.org/wiki/Exponential_distribution). Optional `seed` can be used as a seed for pseudo-random number generator.
See also [rand](#rand) and [rand_normal](#rand_normal).
#### rand_normal
`rand_normal(seed)` is a [transform function](#transform-functions), which returns pseudo-random numbers
with [normal distribution](https://en.wikipedia.org/wiki/Normal_distribution). Optional `seed` can be used as a seed for pseudo-random number generator.
See also [rand](#rand) and [rand_exponential](#rand_exponential).
#### range_avg
`range_avg(q)` is a [transform function](#transform-functions), which calculates the avg value across points per each time series returned by `q`.
#### range_first
`range_first(q)` is a [transform function](#transform-functions), which returns the value for the first point per each time series returned by `q`.
#### range_last
`range_last(q)` is a [transform function](#transform-functions), which returns the value for the last point per each time series returned by `q`.
#### range_linear_regression
`range_linear_regression(q)` is a [transform function](#transform-functions), which calculates [simple linear regression](https://en.wikipedia.org/wiki/Simple_linear_regression)
over the selected time range per each time series returned by `q`. This function is useful for capacity planning and predictions.
#### range_mad
`range_mad(q)` is a [transform function](#transform-functions), which calculates the [median absolute deviation](https://en.wikipedia.org/wiki/Median_absolute_deviation)
across points per each time series returned by `q`.
See also [mad](#mad) and [mad_over_time](#mad_over_time).
#### range_max
`range_max(q)` is a [transform function](#transform-functions), which calculates the max value across points per each time series returned by `q`.
#### range_median
`range_median(q)` is a [transform function](#transform-functions), which calculates the median value across points per each time series returned by `q`.
#### range_min
`range_min(q)` is a [transform function](#transform-functions), which calculates the min value across points per each time series returned by `q`.
#### range_normalize
`range_normalize(q1, ...)` is a [transform function](#transform-functions), which normalizes values for time series returned by `q1, ...` into `[0 ... 1]` range.
This function is useful for correlating time series with distinct value ranges.
See also [share](#share).
#### range_quantile
`range_quantile(phi, q)` is a [transform function](#transform-functions), which returns `phi`-quantile across points per each time series returned by `q`.
`phi` must be in the range `[0...1]`.
#### range_stddev
`range_stddev(q)` is a [transform function](#transform-functions), which calculates [standard deviation](https://en.wikipedia.org/wiki/Standard_deviation)
per each time series returned by `q` on the selected time range.
#### range_stdvar
`range_stdvar(q)` is a [transform function](#transform-functions), which calculates [standard variance](https://en.wikipedia.org/wiki/Variance)
per each time series returned by `q` on the selected time range.
#### range_sum
`range_sum(q)` is a [transform function](#transform-functions), which calculates the sum of points per each time series returned by `q`.
#### range_trim_outliers
`range_trim_outliers(k, q)` is a [transform function](#transform-functions), which drops points located farther than `k*range_mad(q)`
from the `range_median(q)`. E.g. it is equivalent to the following query: `q ifnot (abs(q - range_median(q)) > k*range_mad(q))`.
See also [range_trim_spikes](#range_trim_spikes) and [range_trim_zscore](#range_trim_zscore).
#### range_trim_spikes
`range_trim_spikes(phi, q)` is a [transform function](#transform-functions), which drops `phi` percent of biggest spikes from time series returned by `q`.
The `phi` must be in the range `[0..1]`, where `0` means `0%` and `1` means `100%`.
See also [range_trim_outliers](#range_trim_outliers) and [range_trim_zscore](#range_trim_zscore).
#### range_trim_zscore
`range_trim_zscore(z, q)` is a [transform function](#transform-functions), which drops points located farther than `z*range_stddev(q)`
from the `range_avg(q)`. E.g. it is equivalent to the following query: `q ifnot (abs(q - range_avg(q)) > z*range_avg(q))`.
See also [range_trim_outliers](#range_trim_outliers) and [range_trim_spikes](#range_trim_spikes).
#### range_zscore
`range_zscore(q)` is a [transform function](#transform-functions), which calculates [z-score](https://en.wikipedia.org/wiki/Standard_score)
for points returned by `q`, e.g. it is equivalent to the following query: `(q - range_avg(q)) / range_stddev(q)`.
#### remove_resets
`remove_resets(q)` is a [transform function](#transform-functions), which removes counter resets from time series returned by `q`.
#### round
`round(q, nearest)` is a [transform function](#transform-functions), which rounds every point of every time series returned by `q` to the `nearest` multiple.
If `nearest` is missing then the rounding is performed to the nearest integer.
`ru(free, max)` is a [transform function](#transform-functions), which calculates resource utilization in the range `[0%...100%]` for the given `free` and `max` resources.
For instance, `ru(node_memory_MemFree_bytes, node_memory_MemTotal_bytes)` returns memory utilization over [node_exporter](https://github.com/prometheus/node_exporter) metrics.
#### running_avg
`running_avg(q)` is a [transform function](#transform-functions), which calculates the running avg per each time series returned by `q`.
#### running_max
`running_max(q)` is a [transform function](#transform-functions), which calculates the running max per each time series returned by `q`.
#### running_min
`running_min(q)` is a [transform function](#transform-functions), which calculates the running min per each time series returned by `q`.
#### running_sum
`running_sum(q)` is a [transform function](#transform-functions), which calculates the running sum per each time series returned by `q`.
#### scalar
`scalar(q)` is a [transform function](#transform-functions), which returns `q` if `q` contains only a single time series. Otherwise, it returns nothing.
This function is supported by PromQL.
#### sgn
`sgn(q)` is a [transform function](#transform-functions), which returns `1` if `v>0`, `-1` if `v<0` and `0` if `v==0` for every point `v`
of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
This function is supported by PromQL.
#### sin
`sin(q)` is a [transform function](#transform-functions), which returns `sin(v)` for every `v` point of every time series returned by `q`.
Metric names are stripped from the resulting series. Add [keep_metric_names](#keep_metric_names) modifier in order to keep metric names.
`smooth_exponential(q, sf)` is a [transform function](#transform-functions), which smooths points per each time series returned
by `q` using [exponential moving average](https://en.wikipedia.org/wiki/Moving_average#Exponential_moving_average) with the given smooth factor `sf`.
#### sort
`sort(q)` is a [transform function](#transform-functions), which sorts series in ascending order by the last point in every time series returned by `q`.
`sort_desc(q)` is a [transform function](#transform-functions), which sorts series in descending order by the last point in every time series returned by `q`.
See also [implicit query conversions](#implicit-query-conversions).
The list of supported label manipulation functions:
#### alias
`alias(q, "name")` is [label manipulation function](#label-manipulation-functions), which sets the given `name` to all the time series returned by `q`.
For example, `alias(up, "foobar")` would rename `up` series to `foobar` series.
#### drop_common_labels
`drop_common_labels(q1, ...., qN)` is [label manipulation function](#label-manipulation-functions), which drops common `label="value"` pairs
among time series returned from `q1, ..., qN`.
#### label_copy
`label_copy(q, "src_label1", "dst_label1", ..., "src_labelN", "dst_labelN")` is [label manipulation function](#label-manipulation-functions),
which copies label values from `src_label*` to `dst_label*` for all the time series returned by `q`.
If `src_label` is empty, then the corresponding `dst_label` is left untouched.
#### label_del
`label_del(q, "label1", ..., "labelN")` is [label manipulation function](#label-manipulation-functions), which deletes the given `label*` labels
from all the time series returned by `q`.
#### label_graphite_group
`label_graphite_group(q, groupNum1, ... groupNumN)` is [label manipulation function](#label-manipulation-functions), which replaces metric names
returned from `q` with the given Graphite group values concatenated via `.` char.
For example, `label_graphite_group({__graphite__="foo*.bar.*"}, 0, 2)` would substitute `foo<any_value>.bar.<other_value>` metric names with `foo<any_value>.<other_value>`.
This function is useful for aggregating Graphite metrics with [aggregate functions](#aggregate-functions). For example, the following query would return per-app memory usage:
`labels_equal(q, "label1", "label2", ...)` is [label manipulation function](#label-manipulation-functions), which returns `q` series with identical values for the listed labels
"label1", "label2", etc.
See also [label_match](#label_match) and [label_mismatch](#label_mismatch).
`sort_by_label(q, "label1", ... "labelN")` is [label manipulation function](#label-manipulation-functions), which sorts series in ascending order by the given set of labels.
`sort_by_label_desc(q, "label1", ... "labelN")` is [label manipulation function](#label-manipulation-functions), which sorts series in descending order by the given set of labels.
`sort_by_label_numeric(q, "label1", ... "labelN")` is [label manipulation function](#label-manipulation-functions), which sorts series in ascending order by the given set of labels
`sort_by_label_numeric_desc(q, "label1", ... "labelN")` is [label manipulation function](#label-manipulation-functions), which sorts series in descending order
by the given set of labels using [numeric sort](https://www.gnu.org/software/coreutils/manual/html_node/Version-sort-is-not-the-same-as-numeric-sort.html).
For example, if `foo` series have `bar` label with values `1`, `101`, `15` and `2`, then `sort_by_label_numeric(foo, "bar")`
would return series in the following order of `bar` label values: `101`, `15`, `2` and `1`.
See also [sort_by_label_numeric](#sort_by_label_numeric) and [sort_by_label_desc](#sort_by_label_desc).
### Aggregate functions
**Aggregate functions** calculate aggregates over groups of [rollup results](#rollup-functions).
Additional details:
* By default, a single group is used for aggregation. Multiple independent groups can be set up by specifying grouping labels
in `by` and `without` modifiers. For example, `count(up) by (job)` would group [rollup results](#rollup-functions) by `job` label value
and calculate the [count](#count) aggregate function independently per each group, while `count(up) without (instance)`
would group [rollup results](#rollup-functions) by all the labels except `instance` before calculating [count](#count) aggregate function independently per each group.
Multiple labels can be put in `by` and `without` modifiers.
* If the aggregate function is applied directly to a [series_selector](https://docs.victoriametrics.com/keyConcepts.html#filtering),
then the [default_rollup()](#default_rollup) function is automatically applied before calculating the aggregate.
* Aggregate functions accept arbitrary number of args. For example, `avg(q1, q2, q3)` would return the average values for every point
across time series returned by `q1`, `q2` and `q3`.
* Aggregate functions support optional `limit N` suffix, which can be used for limiting the number of output groups.
For example, `sum(x) by (y) limit 3` limits the number of groups for the aggregation to 3. All the other groups are ignored.
See also [implicit query conversions](#implicit-query-conversions).
The list of supported aggregate functions:
#### any
`any(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns a single series per `group_labels` out of time series returned by `q`.
See also [group](#group).
#### avg
`avg(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns the average value per `group_labels` for time series returned by `q`.
The aggregate is calculated individually per each group of points with the same timestamp.
This function is supported by PromQL.
#### bottomk
`bottomk(k, q)` is [aggregate function](#aggregate-functions), which returns up to `k` points with the smallest values across all the time series returned by `q`.
The aggregate is calculated individually per each group of points with the same timestamp.
`bottomk_avg(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the smallest averages.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `bottomk_avg(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series
with the smallest averages plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [topk_avg](#topk_avg).
#### bottomk_last
`bottomk_last(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the smallest last values.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `bottomk_max(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series
with the smallest maximums plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [topk_last](#topk_last).
#### bottomk_max
`bottomk_max(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the smallest maximums.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `bottomk_max(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series
with the smallest maximums plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [topk_max](#topk_max).
#### bottomk_median
`bottomk_median(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the smallest medians.
If an optional`other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `bottomk_median(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series
with the smallest medians plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [topk_median](#topk_median).
#### bottomk_min
`bottomk_min(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the smallest minimums.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `bottomk_min(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series
with the smallest minimums plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [topk_min](#topk_min).
#### count
`count(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns the number of non-empty points per `group_labels`
for time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
This function is supported by PromQL.
#### count_values
`count_values("label", q)` is [aggregate function](#aggregate-functions), which counts the number of points with the same value
and stores the counts in a time series with an additional `label`, which contains each initial value.
The aggregate is calculated individually per each group of points with the same timestamp.
`distinct(q)` is [aggregate function](#aggregate-functions), which calculates the number of unique values per each group of points with the same timestamp.
per each group of points with the same timestamp. Useful for visualizing big number of time series via a heatmap.
See [this article](https://medium.com/@valyala/improving-histogram-usability-for-prometheus-and-grafana-bc7e5df0e350) for more details.
See also [histogram_over_time](#histogram_over_time) and [histogram_quantile](#histogram_quantile).
#### limitk
`limitk(k, q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns up to `k` time series per each `group_labels`
out of time series returned by `q`. The returned set of time series remain the same across calls.
See also [limit_offset](#limit_offset).
#### mad
`mad(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns the [Median absolute deviation](https://en.wikipedia.org/wiki/Median_absolute_deviation)
per each `group_labels` for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
See also [range_mad](#range_mad), [mad_over_time](#mad_over_time), [outliers_mad](#outliers_mad) and [stddev](#stddev).
#### max
`max(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns the maximum value per each `group_labels`
for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
This function is supported by PromQL.
#### median
`median(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns the median value per each `group_labels`
for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
#### min
`min(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns the minimum value per each `group_labels`
for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
This function is supported by PromQL.
#### mode
`mode(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns [mode](https://en.wikipedia.org/wiki/Mode_(statistics))
per each `group_labels` for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
`outliers_iqr(q)` is [aggregate function](#aggregate-functions), which returns time series from `q` with at least a single point
outside e.g. [Interquartile range outlier bounds](https://en.wikipedia.org/wiki/Interquartile_range) `[q25-1.5*iqr .. q75+1.5*iqr]`
comparing to other time series at the given point, where:
-`iqr` is an [Interquartile range](https://en.wikipedia.org/wiki/Interquartile_range) calculated independently per each point on the graph across `q` series.
-`q25` and `q75` are 25th and 75th [percentiles](https://en.wikipedia.org/wiki/Percentile) calculated independently per each point on the graph across `q` series.
The `outliers_iqr()` is useful for detecting anomalous series in the group of series. For example, `outliers_iqr(temperature) by (country)` returns
per-country series with anomalous outlier values comparing to the rest of per-country series.
See also [outliers_mad](#outliers_mad), [outliersk](#outliersk) and [outlier_iqr_over_time](#outlier_iqr_over_time).
`outliersk(k, q)` is [aggregate function](#aggregate-functions), which returns up to `k` time series with the biggest standard deviation (aka outliers)
`quantiles("phiLabel", phi1, ..., phiN, q)` is [aggregate function](#aggregate-functions), which calculates `phi*`-quantiles for all the time series
returned by `q` and return them in time series with `{phiLabel="phi*"}` label. `phi*` must be in the range `[0...1]`.
The aggregate is calculated individually per each group of points with the same timestamp.
See also [quantile](#quantile).
#### share
`share(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns shares in the range `[0..1]`
for every non-negative points returned by `q` per each timestamp, so the sum of shares per each `group_labels` equals 1.
This function is useful for normalizing [histogram bucket](https://docs.victoriametrics.com/keyConcepts.html#histogram) shares
into `[0..1]` range:
```metricsql
share(
sum(
rate(http_request_duration_seconds_bucket[5m])
) by (le, vmrange)
)
```
See also [range_normalize](#range_normalize).
#### stddev
`stddev(q) by (group_labels)` is [aggregate function](#aggregate-functions), which calculates standard deviation per each `group_labels`
for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
This function is supported by PromQL.
#### stdvar
`stdvar(q) by (group_labels)` is [aggregate function](#aggregate-functions), which calculates standard variance per each `group_labels`
for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
This function is supported by PromQL.
#### sum
`sum(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns the sum per each `group_labels`
for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
This function is supported by PromQL.
#### sum2
`sum2(q) by (group_labels)` is [aggregate function](#aggregate-functions), which calculates the sum of squares per each `group_labels`
for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
#### topk
`topk(k, q)` is [aggregate function](#aggregate-functions), which returns up to `k` points with the biggest values across all the time series returned by `q`.
The aggregate is calculated individually per each group of points with the same timestamp.
`topk_avg(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the biggest averages.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `topk_avg(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series with the biggest averages
plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [bottomk_avg](#bottomk_avg).
#### topk_last
`topk_last(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the biggest last values.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `topk_max(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series with the biggest maximums
plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [bottomk_last](#bottomk_last).
#### topk_max
`topk_max(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the biggest maximums.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `topk_max(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series with the biggest maximums
plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [bottomk_max](#bottomk_max).
#### topk_median
`topk_median(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the biggest medians.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `topk_median(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series with the biggest medians
plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [bottomk_median](#bottomk_median).
#### topk_min
`topk_min(k, q, "other_label=other_value")` is [aggregate function](#aggregate-functions), which returns up to `k` time series from `q` with the biggest minimums.
If an optional `other_label=other_value` arg is set, then the sum of the remaining time series is returned with the given label.
For example, `topk_min(3, sum(process_resident_memory_bytes) by (job), "job=other")` would return up to 3 time series with the biggest minimums
plus a time series with `{job="other"}` label with the sum of the remaining series if any.
See also [bottomk_min](#bottomk_min).
#### zscore
`zscore(q) by (group_labels)` is [aggregate function](#aggregate-functions), which returns [z-score](https://en.wikipedia.org/wiki/Standard_score) values
per each `group_labels` for all the time series returned by `q`. The aggregate is calculated individually per each group of points with the same timestamp.
This function is useful for detecting anomalies in the group of related time series.
MetricsQL supports and extends PromQL subqueries. See [this article](https://valyala.medium.com/prometheus-subqueries-in-victoriametrics-9b1492b720b3) for details.
Any [rollup function](#rollup-functions) for something other than [series selector](https://docs.victoriametrics.com/keyConcepts.html#filtering) form a subquery.
Nested rollup functions can be implicit thanks to the [implicit query conversions](#implicit-query-conversions).
* If lookbehind window in square brackets is missing inside [rollup function](#rollup-functions), then it is automatically set to the following value:
- To `step` value passed to [/api/v1/query_range](https://docs.victoriametrics.com/keyConcepts.html#range-query) or [/api/v1/query](https://docs.victoriametrics.com/keyconcepts/#instant-query)
for all the [rollup functions](#rollup-functions) except of [default_rollup](#default_rollup) and [rate](#rate). This value is known as `$__interval` in Grafana or `1i` in MetricsQL.
For example, `avg_over_time(temperature)` is automatically transformed to `avg_over_time(temperature[1i])`.
- To the `max(step, scrape_interval)`, where `scrape_interval` is the interval between [raw samples](https://docs.victoriametrics.com/keyconcepts/#raw-samples)
for [default_rollup](#default_rollup) and [rate](#rate) functions. This allows avoiding unexpected gaps on the graph when `step` is smaller than `scrape_interval`.
