Struct Histogram 

pub struct Histogram {
    pub edges: Vec<f64>,
    pub counts: Vec<usize>,
}

An equal-width histogram over the non-missing values of a numeric column.

The bin count follows Sturges’ rule (ceil(log2(n) + 1), floored at 1). edges.len() == counts.len() + 1; the final bin is inclusive of its upper edge so the maximum value is never dropped.

Fields

edges: Vec<f64>

Bin edges, ascending. Length is counts.len() + 1 (empty for an empty column).

counts: Vec<usize>

Number of values falling into each bin. Length is one less than edges.

Implementations

impl Histogram

pub fn nbins(&self) -> usize

Returns the number of bins (== counts.len()).

Examples found in repository

examples/numeric_distribution.rs (line 63)

fn main() -> datarust_profile::Result<()> {
    // A synthetic column with a visible right skew: most values cluster low,
    // one extreme value pulls the tail.
    let m = Matrix::from_rows(vec![
        vec![10.0],
        vec![12.0],
        vec![11.0],
        vec![13.0],
        vec![10.0],
        vec![14.0],
        vec![12.0],
        vec![11.0],
        vec![150.0], // outlier — well above the IQR fence
    ])?;

    let p = profile_matrix(&m, Some(&["reaction_ms".into()]))?;
    let col = &p.columns[0];
    let n = col.numeric.as_ref().expect("numeric column");

    println!(
        "Column: {} ({} rows, {} missing)\n",
        col.name, col.count, col.missing_count
    );

    // Central tendency + spread.
    println!("  mean   = {:.2}", n.mean);
    println!("  std    = {:.2}\n", n.std);

    // Five-number summary (min / Q1 / median / Q3 / max).
    println!("  five-number summary:");
    println!("    min    = {:.1}", n.five.min);
    println!("    Q1     = {:.1}", n.five.q1);
    println!("    median = {:.1}", n.five.median);
    println!("    Q3     = {:.1}", n.five.q3);
    println!("    max    = {:.1}\n", n.five.max);

    // Distributional shape (v0.2).
    //   skew ≈ 0  → symmetric; large positive → right tail (our case).
    //   excess kurtosis ≈ 0 → normal-like; positive → heavy-tailed.
    println!("  shape:");
    println!(
        "    skewness  = {:+.3}  ({})",
        n.skewness,
        describe_skew(n.skewness)
    );
    println!(
        "    kurtosis  = {:+.3}  ({})\n",
        n.kurtosis,
        describe_kurtosis(n.kurtosis)
    );

    // Histogram — equal-width bins, Sturges' rule for the count.
    println!("  histogram ({} bins):", n.histogram.nbins());
    let max_count = n.histogram.max_count().max(1);
    for (i, &count) in n.histogram.counts.iter().enumerate() {
        let lo = n.histogram.edges.get(i).copied().unwrap_or(f64::NAN);
        let hi = n.histogram.edges.get(i + 1).copied().unwrap_or(f64::NAN);
        let bar_len = (count as f64 / max_count as f64 * 40.0).round() as usize;
        let bar: String = "█".repeat(bar_len);
        println!("    [{:>7.1}, {:>7.1})  {:>3}  {}", lo, hi, count, bar);
    }
    println!();

    // Outliers — values beyond the Tukey IQR fences.
    println!("  outliers (IQR rule):");
    println!(
        "    {} found ({:.1}% of values)\n",
        n.outlier_count,
        n.outlier_fraction * 100.0
    );

    Ok(())
}

pub fn max_count(&self) -> usize

Returns the count in the tallest bin, or 0 for an empty histogram.

Examples found in repository

examples/numeric_distribution.rs (line 64)

fn main() -> datarust_profile::Result<()> {
    // A synthetic column with a visible right skew: most values cluster low,
    // one extreme value pulls the tail.
    let m = Matrix::from_rows(vec![
        vec![10.0],
        vec![12.0],
        vec![11.0],
        vec![13.0],
        vec![10.0],
        vec![14.0],
        vec![12.0],
        vec![11.0],
        vec![150.0], // outlier — well above the IQR fence
    ])?;

    let p = profile_matrix(&m, Some(&["reaction_ms".into()]))?;
    let col = &p.columns[0];
    let n = col.numeric.as_ref().expect("numeric column");

    println!(
        "Column: {} ({} rows, {} missing)\n",
        col.name, col.count, col.missing_count
    );

    // Central tendency + spread.
    println!("  mean   = {:.2}", n.mean);
    println!("  std    = {:.2}\n", n.std);

    // Five-number summary (min / Q1 / median / Q3 / max).
    println!("  five-number summary:");
    println!("    min    = {:.1}", n.five.min);
    println!("    Q1     = {:.1}", n.five.q1);
    println!("    median = {:.1}", n.five.median);
    println!("    Q3     = {:.1}", n.five.q3);
    println!("    max    = {:.1}\n", n.five.max);

    // Distributional shape (v0.2).
    //   skew ≈ 0  → symmetric; large positive → right tail (our case).
    //   excess kurtosis ≈ 0 → normal-like; positive → heavy-tailed.
    println!("  shape:");
    println!(
        "    skewness  = {:+.3}  ({})",
        n.skewness,
        describe_skew(n.skewness)
    );
    println!(
        "    kurtosis  = {:+.3}  ({})\n",
        n.kurtosis,
        describe_kurtosis(n.kurtosis)
    );

    // Histogram — equal-width bins, Sturges' rule for the count.
    println!("  histogram ({} bins):", n.histogram.nbins());
    let max_count = n.histogram.max_count().max(1);
    for (i, &count) in n.histogram.counts.iter().enumerate() {
        let lo = n.histogram.edges.get(i).copied().unwrap_or(f64::NAN);
        let hi = n.histogram.edges.get(i + 1).copied().unwrap_or(f64::NAN);
        let bar_len = (count as f64 / max_count as f64 * 40.0).round() as usize;
        let bar: String = "█".repeat(bar_len);
        println!("    [{:>7.1}, {:>7.1})  {:>3}  {}", lo, hi, count, bar);
    }
    println!();

    // Outliers — values beyond the Tukey IQR fences.
    println!("  outliers (IQR rule):");
    println!(
        "    {} found ({:.1}% of values)\n",
        n.outlier_count,
        n.outlier_fraction * 100.0
    );

    Ok(())
}

Trait Implementations

impl Clone for Histogram

fn clone(&self) -> Histogram

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for Histogram

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl PartialEq for Histogram

fn eq(&self, other: &Histogram) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for Histogram