chart_bar
Plots events on an bar chart.
chart_bar x=field, y=any, [x_min=any, x_max=any, y_min=any, y_max=any, resolution=duration, fill=any, x_log=bool, y_log=bool, group=any, position=string]
Description
Section titled “Description”Visualizes events with an bar chart on the Tenzir Platform.
x = field
Section titled “x = field”Label for each bar.
y = any
Section titled “y = any”Positions on the y-axis for each data point. Multiple data points for the same group can be be aggregated using an aggregation function.
Multiple y
values and their labels can be specified by using the record
syntax: {name: value, ...}
.
For example, y = {"Avg. Load": mean(load)}
calculates the
mean of the load
field and labels it as Avg. Load
.
x_min = any (optional)
Section titled “x_min = any (optional)”If specified, only charts events where x >= x_min
.
If resolution
is specified, x_min
is floored to create a full bucket.
x_max = any (optional)
Section titled “x_max = any (optional)”If specified, only charts events where x <= x_max
.
If resolution
is specified, x_max
is ceiled to create a full bucket.
y_min = any (optional)
Section titled “y_min = any (optional)”If specified, any y
values less than y_min
will appear clipped out of the
chart.
y_max = any (optional)
Section titled “y_max = any (optional)”If specified, any y
values greater than y_max
will appear clipped out of the
chart.
resolution = duration (optional)
Section titled “resolution = duration (optional)”This option can be specified to create buckets of the given resolution on the
x-axis. An aggregation function must be specified to combine values in the same
bucket when resolution
is specified.
For example, if the resolution is set to 15min
, the x
values are floored
to create buckets of 15 minutes. Any aggregations specified act on that bucket.
fill = any (optional)
Section titled “fill = any (optional)”Optional value to fill gaps and replace null
s with.
x_log = bool (optional)
Section titled “x_log = bool (optional)”If true
, use a logarithmic scale for the x-axis.
Defaults to false
.
y_log = bool (optional)
Section titled “y_log = bool (optional)”If true
, use a logarithmic scale for the y-axis.
Defaults to false
.
group = any (optional)
Section titled “group = any (optional)”Optional expression to group the aggregations with.
position = string (optional)
Section titled “position = string (optional)”Determines how the y
values are displayed. Possible values:
grouped
stacked
Defaults to grouped
.
Examples
Section titled “Examples”Chart count of events imported for every unique schema
Section titled “Chart count of events imported for every unique schema”metrics "import"chart_bar x=schema, y=sum(events), x_min=now()-1d