sample

Returns a random sample from a large dataset, where each data point has an equal probability of being selected. This keyword uses the Bernoulli distribution.

Syntax

sample(percentage-size-of-dataset) repeatable(seed)

Usage

Use sample to project a query on a representative sample from your dataset, where each data point has an equal probability of being selected. sample runs pre-projection.

Add sample and repeatable after the load statement. Any operation performed on the query after the load statement affects only the random sample of data. Let’s look at an example.

q = load "Opportunity" sample(10) repeatable(1);
q = group q by all;
q = foreach q generate count() as 'count';
q = limit q 2000;

Here, the query returns the row count of the sample, 453—around 10% of the dataset's 4.6k rows. The repeatable keyword guarantees that the query always returns the same result. Without the repeatable keyword, the query returns a sample of a slightly different size each time you run it. If you modify your dataset and add more data, then repeatable doesn’t return the same result.

group-by Example

This query returns the counts of opportunities for each stage. Since the query operates on 10% of the dataset, the counts for each stage are approximately 1/10 of the original count.

q = load "Opportunity" sample(10) repeatable(1);
q = group q by 'StageName';
q = foreach q generate 'StageName', count() as 'count';
q = limit q 2000;

filter Example

This query returns only the won opportunities for each stage. Since the query operates on 10% of the dataset, the count for each stage is approximately 1/10 of the original count.

q = load "Opportunity" sample(10);
q = filter q by 'IsWon' == "true";
q = group q by 'StageName';
q = foreach q generate 'StageName', count() as 'count';
q = limit q 2000;

See Also

• Keywords