roundInterval - Round intervals to a lower precision

roundInterval will reduce the precision of an interval value to some desired timeunit. All timeunit fields that are smaller than the desired timeunit will be zeroed out.

Syntax

Like many functions in DataPrime, roundInterval supports two notations, function and method notation. These interchangeable forms allow flexibility in how you structure expressions.

Function notation

roundInterval(interval: interval, scale: timeunit): interval

Method notation

interval: interval.roundInterval(scale: timeunit): interval

Arguments

Name	Type	Required	Description
interval	interval	true	The interval to round
scale	timeunit	true	The highest `timeunit` to render

Example - Grouping time differences into buckets

Consider the following documents:

{
    "time_taken": 2h10m15s
},
{
    "time_taken": 3h8m15s
},
{
    "time_taken": 3h10m15s
},
{
    "time_taken": 53m15s
}

We want to group these values into a bucket per the number of hours taken.

Our first step is to use roundInterval to convert the interval into 1 hour counts.

Function notation

create time_bucket from roundInterval(time_taken, 'h')

Method notation

create time_bucket from time_taken.roundInterval('h')

We now have a new field time_bucket in our document:

{
    "time_taken": 2h10m15s,
    "time_bucket": 2h10m15s
},
{
    "time_taken": 3h8m15s,
    "time_bucket": 3h8m15s
},
{
    "time_taken": 3h10m15s,
    "time_bucket": 3h10m15s
},
{
    "time_taken": 53m15s,
    "time_bucket": 0ns
}

Next, we want to express these intervals as hours.

Function notation

replace time_bucket with formatInterval(time_bucket, 'h')

Method notation

replace time_bucket with time_bucket.formatInterval('h')

This will result in the following document:

{
    "time_taken": 2h10m15s,
    "time_bucket": 2h
},
{
    "time_taken": 5h8m15s,
    "time_bucket": 3h
},
{
    "time_taken": 3h10m15s,
    "time_bucket": 3h
},
{
    "time_taken": 53m15s,
    "time_bucket": 0ns
}

Finally, we can group by time_bucket to count up each value.

groupby time_bucket agg count() as action_count

This results in the following documents:

{
    "time_bucket": 2h,
    "action_count": 1
},
{
    "time_bucket": 3h,
    "action_count": 2
},
{
    "time_bucket": 0ns,
    "action_count": 1
}

parseToTimestamp

Next roundTime