Struct ThetaUnion 

pub struct ThetaUnion { /* private fields */ }

A streaming union accumulator over theta sketches. Accepts any of super::ThetaSketch, CompactThetaSketch, or super::WrappedCompactThetaSketch via the sealed ThetaInput trait.

Implementations

impl ThetaUnion

pub fn update(&mut self, input: &impl ThetaInput)

Merges the given sketch into this union’s running result.

Examples found in repository

examples/bench_theta_update.rs (line 388)

fn bench_union(items: u64, reps: usize) {
    let (a, b) = build_operands(items);
    let mut passes = Vec::with_capacity(reps);
    for _ in 0..reps {
        let start = Instant::now();
        let mut total = 0.0;
        let mut estimate = 0.0;
        for _ in 0..OP_CALLS {
            let mut union = ThetaUnionBuilder::new()
                .lg_k(LG_K)
                .build()
                .expect("fixed valid parameters were rejected");
            union.update(&a);
            union.update(&b);
            estimate = union.get_result(true).get_estimate();
            total += estimate;
        }
        let elapsed = start.elapsed();
        black_box(total);
        passes.push(Pass { elapsed, estimate });
    }
    report_line("union", items, OP_CALLS, &passes, String::new());
}

examples/theta.rs (line 34)

fn main() {
    // Build two sketches representing two overlapping sets of user IDs.
    let mut visitors_day1 = ThetaSketchBuilder::new().lg_k(12).build().unwrap();
    for id in 0..10_000u64 {
        visitors_day1.update_u64(id);
    }

    let mut visitors_day2 = ThetaSketchBuilder::new().lg_k(12).build().unwrap();
    for id in 5_000..15_000u64 {
        visitors_day2.update_u64(id);
    }

    println!(
        "Day 1 unique visitors (estimate): {:.0}",
        visitors_day1.get_estimate()
    );
    println!(
        "Day 2 unique visitors (estimate): {:.0}",
        visitors_day2.get_estimate()
    );

    // Union: total unique visitors across both days.
    let mut union = ThetaUnionBuilder::new().lg_k(12).build().unwrap();
    union.update(&visitors_day1);
    union.update(&visitors_day2);
    let total_unique = union.get_result(true);
    println!(
        "Total unique visitors (union estimate): {:.0}",
        total_unique.get_estimate()
    );

    // Intersection: visitors who came back on day 2.
    let mut intersection = ThetaIntersection::new();
    intersection.update(&visitors_day1);
    intersection.update(&visitors_day2);
    match intersection.get_result(true) {
        Ok(returning) => println!(
            "Returning visitors (intersection estimate): {:.0}",
            returning.get_estimate()
        ),
        Err(e) => println!("No intersection result: {e}"),
    }

    // A-not-b: visitors who only came on day 1.
    let a_not_b = ThetaAnotB::new();
    let day1_only = a_not_b.compute(&visitors_day1, &visitors_day2, true);
    println!(
        "Day-1-only visitors (a-not-b estimate): {:.0}",
        day1_only.get_estimate()
    );

    // Jaccard similarity: how similar are the two days' visitor sets?
    let similarity = jaccard_similarity(&visitors_day1, &visitors_day2);
    println!(
        "Jaccard similarity: {:.3} (range [{:.3}, {:.3}])",
        similarity.estimate, similarity.lower_bound, similarity.upper_bound
    );

    // Serialize a compact sketch for storage/transmission, then restore it.
    let compact = visitors_day1.compact(true);
    let bytes = compact.serialize_compact();
    println!("Serialized day-1 sketch: {} bytes", bytes.len());
    let restored = apache_datasketches::theta::CompactThetaSketch::deserialize(&bytes).unwrap();
    println!("Restored estimate: {:.0}", restored.get_estimate());
}

pub fn get_result(&self, ordered: bool) -> CompactThetaSketch

Returns the union’s current result as a CompactThetaSketch. If ordered is true, the result’s entries are sorted by hash value.

Examples found in repository

examples/bench_theta_update.rs (line 390)

fn bench_union(items: u64, reps: usize) {
    let (a, b) = build_operands(items);
    let mut passes = Vec::with_capacity(reps);
    for _ in 0..reps {
        let start = Instant::now();
        let mut total = 0.0;
        let mut estimate = 0.0;
        for _ in 0..OP_CALLS {
            let mut union = ThetaUnionBuilder::new()
                .lg_k(LG_K)
                .build()
                .expect("fixed valid parameters were rejected");
            union.update(&a);
            union.update(&b);
            estimate = union.get_result(true).get_estimate();
            total += estimate;
        }
        let elapsed = start.elapsed();
        black_box(total);
        passes.push(Pass { elapsed, estimate });
    }
    report_line("union", items, OP_CALLS, &passes, String::new());
}

examples/theta.rs (line 36)

fn main() {
    // Build two sketches representing two overlapping sets of user IDs.
    let mut visitors_day1 = ThetaSketchBuilder::new().lg_k(12).build().unwrap();
    for id in 0..10_000u64 {
        visitors_day1.update_u64(id);
    }

    let mut visitors_day2 = ThetaSketchBuilder::new().lg_k(12).build().unwrap();
    for id in 5_000..15_000u64 {
        visitors_day2.update_u64(id);
    }

    println!(
        "Day 1 unique visitors (estimate): {:.0}",
        visitors_day1.get_estimate()
    );
    println!(
        "Day 2 unique visitors (estimate): {:.0}",
        visitors_day2.get_estimate()
    );

    // Union: total unique visitors across both days.
    let mut union = ThetaUnionBuilder::new().lg_k(12).build().unwrap();
    union.update(&visitors_day1);
    union.update(&visitors_day2);
    let total_unique = union.get_result(true);
    println!(
        "Total unique visitors (union estimate): {:.0}",
        total_unique.get_estimate()
    );

    // Intersection: visitors who came back on day 2.
    let mut intersection = ThetaIntersection::new();
    intersection.update(&visitors_day1);
    intersection.update(&visitors_day2);
    match intersection.get_result(true) {
        Ok(returning) => println!(
            "Returning visitors (intersection estimate): {:.0}",
            returning.get_estimate()
        ),
        Err(e) => println!("No intersection result: {e}"),
    }

    // A-not-b: visitors who only came on day 1.
    let a_not_b = ThetaAnotB::new();
    let day1_only = a_not_b.compute(&visitors_day1, &visitors_day2, true);
    println!(
        "Day-1-only visitors (a-not-b estimate): {:.0}",
        day1_only.get_estimate()
    );

    // Jaccard similarity: how similar are the two days' visitor sets?
    let similarity = jaccard_similarity(&visitors_day1, &visitors_day2);
    println!(
        "Jaccard similarity: {:.3} (range [{:.3}, {:.3}])",
        similarity.estimate, similarity.lower_bound, similarity.upper_bound
    );

    // Serialize a compact sketch for storage/transmission, then restore it.
    let compact = visitors_day1.compact(true);
    let bytes = compact.serialize_compact();
    println!("Serialized day-1 sketch: {} bytes", bytes.len());
    let restored = apache_datasketches::theta::CompactThetaSketch::deserialize(&bytes).unwrap();
    println!("Restored estimate: {:.0}", restored.get_estimate());
}

pub fn reset(&mut self)

Resets this union to its initial, empty state.

Trait Implementations

impl Send for ThetaUnion