Struct Table 

pub struct Table { /* private fields */ }

A small, owned, rectangular table-like data set.

External guarantees: row order is preserved; column order is preserved; column names are stable after loading. Operations return new owned Table values; no borrowed view lifetimes appear in normal use.

Implementations

impl Table

pub fn from_csv_str(input: &str) -> Result<Table, MattenDataError>

Parse a Table from a CSV string.

The first row is the header. Empty input, empty or duplicate header names, and ragged rows are reported as MattenDataError (never a panic).

use matten_data::Table;
let table = Table::from_csv_str("a,b\n1,2\n3,4").unwrap();
assert_eq!(table.row_count(), 2);
assert_eq!(table.column_names(), &["a".to_string(), "b".to_string()]);

Examples found in repository

examples/csv_to_tensor.rs (line 15)

fn main() -> Result<(), matten_data::MattenDataError> {
    // A small, messy table: mixed types and a missing cell.
    let csv = "\
region,sales,cost,quantity
north,100,40,5
south,150,,7
east,120,55,6";

    let table = Table::from_csv_str(csv)?;

    // Inspect what we have before converting anything.
    println!("{}", table.schema_summary());

    // Select only the numeric columns we want, fill the one missing cost with 0,
    // convert explicitly, and produce a [rows, columns] f64 tensor.
    let tensor = table
        .select_columns(["sales", "cost", "quantity"])?
        .fill_missing(0.0)?
        .try_numeric()?
        .to_tensor()?;

    println!("tensor shape: {:?}", tensor.shape());
    println!("tensor data : {:?}", tensor.as_slice());

    assert_eq!(tensor.shape(), &[3, 3]);
    Ok(())
}

pub fn from_csv_path<P: AsRef<Path>>(path: P) -> Result<Table, MattenDataError>

Parse a Table from a CSV file at path.

I/O failures (for example a missing file) are reported as MattenDataError::Io with the path and underlying error preserved.

impl Table

pub fn row_count(&self) -> usize

Number of data rows.

pub fn column_count(&self) -> usize

Number of columns.

pub fn column_names(&self) -> &[String]

Column names, in column order.

pub fn schema_summary(&self) -> SchemaSummary

A small, displayable schema summary (row/column counts, per-column missing counts and inferred kinds). Does not perform expensive analysis.

Examples found in repository

examples/csv_to_tensor.rs (line 18)

fn main() -> Result<(), matten_data::MattenDataError> {
    // A small, messy table: mixed types and a missing cell.
    let csv = "\
region,sales,cost,quantity
north,100,40,5
south,150,,7
east,120,55,6";

    let table = Table::from_csv_str(csv)?;

    // Inspect what we have before converting anything.
    println!("{}", table.schema_summary());

    // Select only the numeric columns we want, fill the one missing cost with 0,
    // convert explicitly, and produce a [rows, columns] f64 tensor.
    let tensor = table
        .select_columns(["sales", "cost", "quantity"])?
        .fill_missing(0.0)?
        .try_numeric()?
        .to_tensor()?;

    println!("tensor shape: {:?}", tensor.shape());
    println!("tensor data : {:?}", tensor.as_slice());

    assert_eq!(tensor.shape(), &[3, 3]);
    Ok(())
}

pub fn select_columns<I, S>(&self, columns: I) -> Result<Table, MattenDataError>

where I: IntoIterator<Item = S>, S: AsRef<str>,

Select columns by name, returning a new Table.

Behavior (RFC-034 §5.3): preserves the requested column order; errors with MattenDataError::MissingColumn if a requested column does not exist; rejects duplicate selections with MattenDataError::DuplicateSelection; an empty selection is MattenDataError::EmptySelection.

Examples found in repository

examples/csv_to_tensor.rs (line 23)

fn main() -> Result<(), matten_data::MattenDataError> {
    // A small, messy table: mixed types and a missing cell.
    let csv = "\
region,sales,cost,quantity
north,100,40,5
south,150,,7
east,120,55,6";

    let table = Table::from_csv_str(csv)?;

    // Inspect what we have before converting anything.
    println!("{}", table.schema_summary());

    // Select only the numeric columns we want, fill the one missing cost with 0,
    // convert explicitly, and produce a [rows, columns] f64 tensor.
    let tensor = table
        .select_columns(["sales", "cost", "quantity"])?
        .fill_missing(0.0)?
        .try_numeric()?
        .to_tensor()?;

    println!("tensor shape: {:?}", tensor.shape());
    println!("tensor data : {:?}", tensor.as_slice());

    assert_eq!(tensor.shape(), &[3, 3]);
    Ok(())
}

pub fn fill_missing( &self, value: impl Into<CellValue>, ) -> Result<Table, MattenDataError>

Fill every missing cell with value, returning a new Table.

Missing values are never silently turned into zero; filling is always explicit (RFC-035 §6). Non-missing cells and the shape are unchanged.

Examples found in repository

examples/csv_to_tensor.rs (line 24)

fn main() -> Result<(), matten_data::MattenDataError> {
    // A small, messy table: mixed types and a missing cell.
    let csv = "\
region,sales,cost,quantity
north,100,40,5
south,150,,7
east,120,55,6";

    let table = Table::from_csv_str(csv)?;

    // Inspect what we have before converting anything.
    println!("{}", table.schema_summary());

    // Select only the numeric columns we want, fill the one missing cost with 0,
    // convert explicitly, and produce a [rows, columns] f64 tensor.
    let tensor = table
        .select_columns(["sales", "cost", "quantity"])?
        .fill_missing(0.0)?
        .try_numeric()?
        .to_tensor()?;

    println!("tensor shape: {:?}", tensor.shape());
    println!("tensor data : {:?}", tensor.as_slice());

    assert_eq!(tensor.shape(), &[3, 3]);
    Ok(())
}

pub fn try_numeric(&self) -> Result<NumericTable, MattenDataError>

Convert the table to an explicit numeric table (RFC-035 §7).

Strict conversion: Int/Float become f64; Bool and Text are rejected (MattenDataError::NonNumericValue); a remaining Missing cell is rejected (MattenDataError::MissingValue). Text is never parsed as a number by default.

Examples found in repository

examples/csv_to_tensor.rs (line 25)

fn main() -> Result<(), matten_data::MattenDataError> {
    // A small, messy table: mixed types and a missing cell.
    let csv = "\
region,sales,cost,quantity
north,100,40,5
south,150,,7
east,120,55,6";

    let table = Table::from_csv_str(csv)?;

    // Inspect what we have before converting anything.
    println!("{}", table.schema_summary());

    // Select only the numeric columns we want, fill the one missing cost with 0,
    // convert explicitly, and produce a [rows, columns] f64 tensor.
    let tensor = table
        .select_columns(["sales", "cost", "quantity"])?
        .fill_missing(0.0)?
        .try_numeric()?
        .to_tensor()?;

    println!("tensor shape: {:?}", tensor.shape());
    println!("tensor data : {:?}", tensor.as_slice());

    assert_eq!(tensor.shape(), &[3, 3]);
    Ok(())
}

Trait Implementations

impl Clone for Table

fn clone(&self) -> Table

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Table

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

Formats the value using the given formatter.